CN113298912A

CN113298912A - Commodity picture processing method and device and server

Info

Publication number: CN113298912A
Application number: CN202010337230.2A
Authority: CN
Inventors: 韩玉洁; 葛妮瑜; 黄非; 刘婕; 巩媛媛; 宋蒙蒙; 王鹏杰; 葛正晗; 施杨斌; 郑莹; 周婷; 聂一兵; 樊楷; 夏明泽; 白鑫; 徐磊; 林晓雯
Original assignee: Alibaba Group Holding Ltd
Current assignee: Alibaba Singapore Holdings Pte Ltd
Priority date: 2020-04-26
Filing date: 2020-04-26
Publication date: 2021-08-24

Abstract

The specification provides a commodity picture processing method, a commodity picture processing device and a server. In one embodiment, the commodity picture processing method performs preset processing matched with a target electronic market scene on a target picture to be processed first, so as to acquire target text data matched with the target electronic market scene from the target picture; then, the target text data is translated into a target language in a targeted manner to obtain translated target text data; and then, the image combination processing can be carried out according to the translated target text data and the target picture to obtain the commodity picture suitable for the target electronic market scene. Therefore, the text data in the target picture can be accurately translated into the target language, and meanwhile, the commodity picture which is suitable for the target e-commerce scene and has a good putting effect is obtained.

Description

Commodity picture processing method and device and server

Technical Field

The specification belongs to the technical field of internet, and particularly relates to a commodity picture processing method, a commodity picture processing device and a server.

Background

When a merchant puts a commodity sold on a shopping website of country a into a shopping website of country B for sale, text data on a commodity picture used when the commodity is sold on the shopping website of country a is often translated into a language used by country B, and then the translated commodity picture is put on a shopping website page of country B.

At present, there is a need for a picture processing method capable of translating text data in a commodity picture into a target language for a specific target e-commerce scene and obtaining a commodity picture suitable for the target e-commerce scene and having a good delivery effect.

Disclosure of Invention

The specification provides a commodity picture processing method, a commodity picture processing device and a commodity picture server, so that a commodity picture which is suitable for a target e-commerce scene and has a good putting effect can be obtained while text data in the target picture is accurately translated into a target language aiming at the target e-commerce scene.

The commodity picture processing method, device and server provided by the specification are realized as follows:

a commodity picture processing method comprises the following steps: acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to commodities and text data related to the commodities; presetting the target picture to acquire target text data matched with a target electronic market using a target language from the target picture; translating the target text data into a target language to obtain translated target text data; and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

A picture processing method comprises the following steps: acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to an object and text data related to the object; presetting the target picture to acquire target text data matched with a target application scene from the target picture; translating the target text data into a target language to obtain translated target text data; and according to the translated target text data and the target picture, carrying out image combination processing to obtain a processed picture suitable for a target scene.

A commodity picture processing method comprises the following steps: displaying a picture processing request interface, and receiving a target picture and an indication parameter of a target language through the picture processing request interface, wherein the target picture comprises image data related to a commodity and text data related to the commodity; generating and sending a picture processing request to a server according to the target picture and the indication parameters of the target language, wherein the server is used for carrying out preset processing on the target picture so as to obtain target text data matched with the target electronic market scene using the target language from the target picture; translating the target text data into a target language to obtain translated target text data; according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene; and receiving and displaying the commodity picture.

A commodity picture processing method comprises the following steps: acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to commodities and text data related to the commodities; obtaining a commodity picture suitable for a target electronic market scene according to the target picture and the indication parameters of the target language, wherein the commodity picture suitable for the target electronic market scene comprises the steps of translating target text data matched with the target electronic market scene in the target picture into the target language, and processing the obtained commodity picture through image combination; and receiving and replacing the target picture in the target E-commerce scene by using the commodity picture suitable for the target E-commerce scene according to the issuing instruction fed back by the user aiming at the commodity picture suitable for the target E-commerce scene.

A picture processing method comprises the following steps: receiving and displaying a target picture sent by a second user to a first user; acquiring an indication parameter of a target language of a first user for the target picture; generating and sending a picture processing request aiming at the target picture to a server according to the indication parameter of the target language, wherein the server is used for responding to the picture processing request and carrying out preset processing on the target picture so as to obtain target text data from the target picture; translating the target text data into a target language to obtain translated target text data; carrying out image combination processing according to the translated target text data and the target image to obtain a processed image suitable for a first user; and acquiring and displaying the processed picture to the first user.

A picture processing method comprises the following steps: acquiring a target picture, wherein the target picture is a picture which is indicated by a second user and sent to a first user; determining whether the language in the target picture matches a first user; under the condition that the language in the target picture is determined to be not matched with the first user, determining the language used by the first user as the target language, and processing the target picture based on the target language to obtain a processed picture suitable for the first user; and sending the processed picture to the first user.

A device for processing a picture of an article, comprising: the system comprises a receiving module, a processing module and a display module, wherein the receiving module is used for acquiring a target picture and an indication parameter of a target language, and the target picture comprises image data related to commodities and text data related to the commodities; the preset processing module is used for carrying out preset processing on the target picture so as to acquire target text data matched with a target electronic market using a target language from the target picture; the translation processing module is used for translating the target text data into a target language to obtain translated target text data; and the image combination processing module is used for carrying out image combination processing according to the translated target text data and the target image to obtain a commodity image suitable for the target electronic market scene.

A server comprises a processor and a memory for storing processor-executable instructions, wherein the processor executes the instructions to acquire a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to commodities and text data related to the commodities; presetting the target picture to acquire target text data matched with a target electronic market using a target language from the target picture; translating the target text data into a target language to obtain translated target text data; and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

A terminal device comprises a processor and a memory for storing processor executable instructions, wherein the processor realizes a display picture processing request interface when executing the instructions, and receives a target picture and an indication parameter of a target language through the picture processing request interface, wherein the target picture comprises image data related to commodities and text data related to the commodities; generating and sending a picture processing request to a server according to the target picture and the indication parameters of the target language, wherein the server is used for carrying out preset processing on the target picture so as to obtain target text data matched with the target electronic market scene using the target language from the target picture; translating the target text data into a target language to obtain translated target text data; according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene; and receiving and displaying the commodity picture.

A computer readable storage medium having stored thereon computer instructions that, when executed, enable obtaining a target picture and an indication parameter of a target language, wherein the target picture contains image data related to a commodity and text data related to the commodity; presetting the target picture to acquire target text data matched with a target electronic market using a target language from the target picture; translating the target text data into a target language to obtain translated target text data; and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

According to the processing method, device and server for the commodity pictures, preset processing matched with the target E-market scene is performed on a target picture to be translated into a target language at first, so that target text data matched with the target E-market scene is obtained from the target picture; then, the target text data is translated into a target language in a targeted manner to obtain translated target text data; and then, the image combination processing can be carried out according to the translated target text data and the target picture to obtain the commodity picture suitable for the target electronic market scene. Therefore, the text data in the target picture can be accurately translated into the target language, and meanwhile, the commodity picture which is suitable for the target e-commerce scene and has a good putting effect is obtained.

Drawings

In order to more clearly illustrate the embodiments of the present specification, the drawings needed to be used in the embodiments will be briefly described below, and the drawings in the following description are only some of the embodiments described in the present specification, and it is obvious to those skilled in the art that other drawings can be obtained according to the drawings without any creative effort.

Fig. 1 is a schematic diagram of an embodiment of a system structure composition to which a method for processing a commodity picture provided by an embodiment of the present specification is applied;

fig. 2 is a schematic diagram of an embodiment of a method for processing a commodity picture provided by an embodiment of the present specification, in a scene example;

fig. 3 is a schematic diagram of an embodiment of a method for processing a commodity picture provided by an embodiment of the present specification, in a scene example;

fig. 4 is a schematic diagram of an embodiment of a method for processing a commodity picture provided by an embodiment of the present specification, in a scene example;

fig. 5 is a schematic diagram of an embodiment of a method for processing a commodity picture provided by an embodiment of the present specification, in a scene example;

fig. 6 is a schematic diagram of an embodiment of a method for processing a commodity picture provided by an embodiment of the present specification, in a scene example;

fig. 7 is a schematic diagram of an embodiment of a method for processing a commodity picture provided by an embodiment of the present specification, in a scene example;

fig. 8 is a schematic diagram of an embodiment of a method for processing a commodity picture provided by an embodiment of the present specification, in a scene example;

fig. 9 is a schematic diagram of an embodiment of a method for processing a commodity picture provided by an embodiment of the present specification, in a scene example;

fig. 10 is a flowchart illustrating a method for processing a commodity picture according to an embodiment of the present disclosure;

FIG. 11 is a schematic structural component diagram of a server provided in an embodiment of the present description;

fig. 12 is a schematic structural diagram of a product picture processing device according to an embodiment of the present specification.

Detailed Description

In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present specification, and not all of the embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments in the present specification without any inventive step should fall within the scope of protection of the present specification.

The embodiment of the specification provides a method for processing a commodity picture, which can be particularly applied to a processing system comprising a server and a terminal device. Specifically, as shown in fig. 1, the server and the terminal device may be connected in a wired or wireless manner for data interaction.

Specifically, the user can input a target picture to be translated through the terminal device, and set an indication parameter of a target language desired to be translated. And the terminal equipment responds to the operation of the user, and generates and sends a corresponding picture processing request to the server. The image processing request at least carries a target image and an indication parameter of a target language, and the target image comprises image data related to commodities and text data related to the commodities.

The server receives a picture processing request; presetting the target picture to acquire target text data matched with a target electronic market using a target language from the target picture; translating the target text data into a target language to obtain translated target text data; according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene; and sending the commodity picture to the terminal equipment.

And the terminal equipment receives the commodity picture sent by the server and displays the commodity picture to the user.

In this embodiment, the server may specifically include a server that is applied to a network platform side and is in charge of data processing in a background, and is capable of implementing functions such as data transmission and data processing. Specifically, the server may be, for example, an electronic device having data operation, storage function and network interaction function. Alternatively, the server may be a software program running in the electronic device and providing support for data processing, storage and network interaction. In the present embodiment, the number of servers is not particularly limited. The server may specifically be one server, or may also be several servers, or a server cluster formed by several servers.

In this embodiment, the terminal device may specifically include a front-end device that is applied to a user side and can implement functions such as data acquisition and data transmission. Specifically, the terminal device may be, for example, a desktop computer, a tablet computer, a notebook computer, a smart phone, a digital assistant, a smart wearable device, and the like. Alternatively, the terminal device may be a software application capable of running in the electronic device. For example, it may be some APP running on a cell phone, etc.

In a specific scenario example, referring to fig. 2, by applying the method for processing a product picture provided by the embodiment of the present specification, a merchant a can quickly and conveniently translate a product picture in a detail page of a product No. 1 sold on a shopping website of country a, which is sold by the merchant a, into a product picture of the product No. 1 in language of country B, and is suitable for displaying and releasing the product picture of the product No. 1 on the shopping website of country B, and publish the product picture of the product No. 1 on the shopping website of country B, so as to attract customers of country B to purchase, and improve the yield of the shopping website of country B.

In this scenario example, merchant A has previously sold item number 1 (e.g., TL brand coat) on the shopping website in nation A. As shown in fig. 3, in order to better show the advantages and features of the product No. 1 to the customers, more customers are attracted to pay attention. The merchant a publishes a product picture (which may be abbreviated as picture 1) containing image data related to the product No. 1 and text data related to the content such as the related introduction and promotion of the product No. 1 on the detail page of the product No. 1 in a shop of a shopping site in country a.

Referring to fig. 3, since the picture 1 is oriented to the e-commerce scenario of country a, the text data in fig. 1 uses the language of country a: chinese language. In order to attract the browsing of the client in the country A and achieve better popularization and release effects, the merchant A also carries out corresponding editing processing on the text data used in the picture 1 by combining the language culture characteristics of the country A, so that the advantages and characteristics of the product No. 1 and the text contents such as discount information and the like of the merchant A aiming at the product No. 1 recently can be clearly described for the client through the text data, and the interest of the client can be psychologically aroused through the text data. In addition, the merchant A also adopts an e-commerce scene suitable for the nation A, and lays specific text data in the picture 1 according to a typesetting rule easily accepted by the nation A customers. For example, in order to attract the attention of customers and highlight the characteristics of the product No. 1, and to better match the reading habits of customers in country a, the merchant a will describe two text data of the main selling points "good quality" and "comfortable and breathable" of the product No. 1, and the two text data are arranged in the vertical arrangement manner in fig. 1. By the typesetting layout, the putting effect of the picture 1 can be further improved, more customers are attracted to pay attention to the commodity No. 1, and the commodity is interested.

Currently, merchant A wants to promote and sell product No. 1 on the shopping website of B nation. And the language used in nation B is english. Therefore, the merchant a usually needs to translate the text data in the picture 1 into english, and then release the translated picture 1 to a detailed description page about the product No. 1 on the B country shopping website for display to the B country user.

However, it is considered that since the difference between the client in country B and the client in country a is large in terms of language, culture, and the like, the appearance and the psychological feeling of the text data and the image data in the same commodity picture and the layout of the two data also have a large difference, so that the e-commerce scene in country B and the e-commerce scene in country a have a large difference.

In this case, if the first merchant simply translates all the text data appearing in picture 1 into english through machine translation, and then puts the translated text data on the shopping site of country B. The translated picture 1 obtained in this way is not suitable for the e-commerce scene of the country B, and after browsing the translated picture 1, the customer of the country B is not attracted by the commodity picture like the customer of the country a, which affects the putting effect of the picture 1. Meanwhile, the translated picture 1 obtained in the above manner is relatively easy to be translated inaccurately, or a part of text data (for example, a trademark of a commodity or the like) which should not be translated in the commodity picture is translated incorrectly, which further affects the putting effect of the picture 1.

In the example of the scenario, the merchant a itself is not familiar with the language of the country B, the cultural characteristics of the country B, the language habit, and the like, and at this time, the merchant a may use its own computer as a terminal device to interact with the server responsible for the commodity picture processing, so as to perform corresponding processing on the picture 1 used on the shopping website of the country a, obtain the commodity picture translated into english and suitable for the e-commerce scenario of the country B.

Specifically, the merchant a may input the picture 1 in a picture processing request interface displayed by the terminal device, and input "english" as an indication parameter of the target language in a target language input field of the interface. The terminal device can respond to the operation, acquire the picture 1 and the indication parameter of the target language, and generate a picture processing request for the picture 1 according to the data. The picture processing request at least carries the picture 1 and the indication parameter "english" of the target language. And the terminal equipment sends the picture processing request to the server in a wired or wireless mode.

The server receives the picture processing request, and obtains a picture 1 to be processed and a target language indication parameter from the picture processing request through data analysis: english. The server can determine that the target e-commerce scene to which the picture 1 is directed is a B-country e-commerce scene using english according to the target language indication parameter. And then the server can select a processing mode matched with the B country electric market scene, and the preset processing is firstly carried out on the picture 1 so as to extract target text data which meets the requirements and needs to be translated and is matched with the B country electric market scene from the picture 1.

In a specific implementation, the server may first perform OCR (optical character recognition) on the picture 1, and recognize text data from the picture 1, and data such as position coordinates of the text data in the picture 1 as an OCR recognition result.

Of course, the above-listed OCR recognition results are only an illustrative description. In specific implementation, according to specific situations and processing needs, a background color of the text data in the picture 1, character features of the text data (for example, a size of a character, an arrangement direction of the character, a color of the character, and the like), a background pattern of the text data in the picture 1, an association relationship between the text data, and the like may be further recognized from the picture 1 through OCR recognition as an OCR recognition result.

In this scenario example, the server may select a preset processing rule matching the B country e-market scene to perform specific processing on the OCR recognition result obtained by the recognition, so as to screen out target text data which matches the B country e-market scene and needs to be translated into english.

The preset processing rules may specifically include processing rules determined by learning languages, cultures, reading habits related to various e-commerce scenes, and commodity pictures, marketing strategies, popularization phrases and the like used by existing commodities in the e-commerce scene in advance.

Specifically, the preset processing rule may be determined for an e-commerce scene of a specific area, based on local cultural characteristics (for example, language characteristics used in the area, and a long-term moral or abstinence formed in the area) of the area corresponding to the e-commerce scene, and characteristics of information content, shopping habits, impression and the like of a consumer in the area paying attention to shopping on a shopping website.

In the present scenario example, as shown in fig. 4, it is considered that in most e-commerce scenarios, a consumer often relatively wants text data (for example, characters printed on a product, or characters printed on a product package, etc.) set on a product body reflected in a product picture in a detail page on a shopping website, and text data on a product label carried by the product, etc. to be completely consistent with a product real object purchased by the consumer. Therefore, in an e-commerce scenario, text data on a product body, text data on a product label, and the like are generally required to be not translated. In addition, many commodity pictures also contain the trademark patterns of the commodities. Some brands have patterns that themselves contain text data. However, the text data contained in the trademark in the e-commerce scene is often just a kind of logo for distinguishing other brands of goods. Therefore, it is also required that no translation process be performed on the text data included in the trademark in the e-commerce scenario.

Based on the processing requirements in the e-commerce scene, when the server specifically processes the OCR recognition result according to a preset processing rule, the server may determine text features of each text data in the OCR recognition result. For example, it is determined whether the following text features exist in each text data in the OCR recognition result: whether text data contained in the trademark of the commodity exists, whether patterns corresponding to the trademark of the commodity are contained in the text data, whether the content and the form represented by the text data are consistent with the content and the form of the commodity label, and the like. Further, according to the determined text characteristics, text data which accords with the first type of text characteristics (for example, text data contained in a trademark exists in the text data; and/or patterns contained in the trademark exist in the text data; and/or content and form represented by the text data are consistent with content and form of a commodity label, and the like) in the OCR recognition result can be screened out as the first type of text data. And further, the first type of text data can be determined to be text data which does not need to be subjected to subsequent translation processing, and the first type of text data is removed from the OCR recognition result. And then, acquiring target text data needing translation processing according to the OCR recognition result without the first type of text data.

The first type of text data may specifically include text data set on a commodity body in a commodity picture, and/or text data set on a trademark of a commodity. In an e-commerce scenario, no translation processing is typically required for the first type of text data described above.

Of course, in specific implementation, the image features related to the text data may be determined according to the OCR recognition result. For example, a background pattern of text data in a picture of a commodity, and the like. And screening out the first type of text data from the OCR recognition result according to the image characteristics related to the text data. For example, text data with the background pattern being the commodity body pattern can be screened from the OCR recognition result according to the background pattern of the text data in the commodity picture as the first type of text data. In addition, text data with a background pattern having a high similarity to the trademark pattern of the commodity may be screened from the OCR recognition result according to the background pattern of the text data in the commodity picture, and the text data may be used as the first type of text data. Of course, the manner in which the first type of text data is determined as listed above is merely illustrative. In specific implementation, according to specific conditions of the commodity, the first type of text data can be screened from the OCR recognition result in other modes. The present specification is not limited to these.

In this scenario example, it is considered that when a merchant makes an edited commodity picture, in order to make the layout of text data in the picture beautiful, a long text data segment is often split into multiple lines and arranged in the picture 1. For example, as shown in FIG. 5. The length of the text data of each line in the figure is approximately the same. However, the text content represented by a single line of text data is often incomplete, i.e. the text data of a single line is not a complete text data. In general, when OCR recognition is performed, recognition processing is often performed on text data on a line-by-line basis, and when translation is subsequently performed, translation processing is also performed on text data on a line-by-line basis. For example, a single line of text data may be treated as one text data and machine translated separately. At this time, because the line of text data itself is not complete text data, translation errors are easy to occur during translation, so that the original text content cannot be accurately expressed by the translated text data.

Meanwhile, because the languages used are different, if the text data with the same length of the plurality of lines are directly and respectively translated, the lengths of the obtained corresponding translated text data of the plurality of lines are also easy to be different, which can be seen from fig. 5. When the images are backfilled, if the translated multi-line text data is continuously and directly backfilled, the layout and the typesetting of the text data in the processed commodity images are correspondingly disordered, and the sensory experience of consumption is influenced.

Based on the above situation, when the server performs specific processing on the OCR recognition result according to the preset processing rule, the server may also merge multiple lines of text data describing the same segment of content in the OCR recognition result into an integral text data through multiple line merging first, and use the integral text data as a target text data, so as to perform subsequent translation processing and image merging processing.

Specifically, referring to fig. 6, the server may find, according to the position coordinates of the text data in the OCR recognition result, multiple lines of text data with adjacent position coordinates, the same character characteristics and the same length as the multiple lines of text data belonging to the same text region and representing the same segment of content. Further, the text data of the plurality of lines may be spliced together in the order from top to bottom and from left to right, and the plurality of lines may be combined into a piece of text data as a total text data. And then, the integrated text data can be directly subjected to translation processing and image combination processing.

For example, referring to fig. 6, when the subsequent graph combination processing is specifically performed, the translated total text data may be further split into multiple lines of text data with the same length, and the split multiple lines of text data may be arranged in a graph in a backfill manner.

In the scene example, in order to attract the attention of the consumer and achieve a good putting effect, specific characters in the text data in the commodity picture are not always arranged in a conventional arrangement direction or a conventional arrangement mode. For example, as shown in fig. 5, in order to highlight the selling point of item No. 1, the characters in the two text data "good quality", "comfortable and breathable" in the figure are vertically arranged. For such a situation, if the text data is recognized directly according to the OCR for translation processing, and then the translated text data is directly utilized for backfill and drawing, the layout effect in the original commodity picture often cannot be restored. Moreover, since the text data before and after translation are in different languages, the reading habits of the included characters are also different, so that the arrangement mode or the arrangement sequence of the characters in the text data before translation is not necessarily suitable for the characters in the text data after translation.

Based on the above situation, when the server specifically processes the OCR recognition result according to the preset processing rule, the server may further determine the arrangement direction of the characters according to the character features of the text data in the OCR recognition result. When the arrangement direction of characters in the partial text data in the OCR result is not the conventional horizontal arrangement, for example, the vertical arrangement, the text data with the vertical arrangement of the characters can be screened out from the OCR recognition result as the second type text data. Further, for the convenience of subsequent translation processing, the second type text data may be converted into text data arranged along a preset character arrangement direction (for example, a conventional character arrangement direction: horizontal arrangement) according to the character arrangement direction, so as to obtain processed second type text data, which is used as a target text data for subsequent translation processing; and meanwhile, vertical arrangement labels are also arranged in the processed second-class text data, so that the translated second-class text data can be arranged according to the original arrangement direction when the graph is filled back subsequently.

The second type of text data may be specifically understood as text data in which character features in the commodity picture are different from conventional preset character features. Specifically, the text data in which the character features of the included characters are vertically arranged, or the text data in which the character features of the included characters are radially arranged, or the text data in which the character features of the included characters are larger than a preset threshold value, may be included.

Of course, the second type of text data listed above is only an illustrative illustration. In specific implementation, after the second type of text data is screened out, adjustment related to character features can be performed on the second type of text data, the second type of text data is converted into processed second type of text data suitable for subsequent translation processing, and a data tag corresponding to the character features of characters included in the text data is set on the processed second type of text data, so that during subsequent backfilling and matching, the typesetting and layout effects of the second type of text data can be restored more accurately in a targeted manner according to the data tag.

In the present scenario example, it is also considered that metering descriptions associated with the goods are often involved in e-commerce scenarios. For example, a description of the size of the clothing, a description of the shoe size of the shoes, a description of the weight of the snacks, etc. In different areas, under the influence of different cultures, when the same type of measurement of the same commodity is described, the adopted measurement units are different. For example, for the same product weight description, the unit of measurement commonly used in nation a is "jin", and the unit of measurement used in nation B is "Kg". In this case, if the text data (denoted as the third type text data) including the unit word is directly translated into the target language without being processed, the obtained translated text data is often relatively poor in effect. For example, if the third type of text data "4 jin" containing unit words in the commodity picture applicable to country a is directly translated into the target language: "four jin". Although the translated text data itself has no errors, the weight unit commonly used in daily life is usually kilogram for the consumers in country B, and the weight unit of "jin" is strange to the consumers and is not easy to understand. Therefore, when the user sees the "fourjin" displayed in the commodity picture, the user cannot understand the specific weight represented by the part of the text data, and the user is troubled by the consumer in country B, and the putting effect of the commodity picture is affected.

Based on the above situation, when the server performs specific processing on the OCR recognition result according to a preset processing rule, the server may further perform unit word search on the OCR recognition result, and screen out text data containing unit words from the OCR recognition result as third-type text data. Further, it is determined whether the unit word used in the third type of text data matches the E-market scene of nation B. Under the condition that the unit words in the third type of text data are determined to be not matched with the target E-market scenery, determining matched unit words corresponding to the unit words; and modifying the unit words in the third type of text data and the numerical characters adjacent to the unit words according to the conversion rule of the matched unit words to obtain the processed third type of text data which is used as target text data.

For example, when it is determined that the unit word "jin" used in the third type of text data "weight 4 jin" is not a unit word commonly used in the e-commerce scene of country B, it may be determined that the unit word in the third type of text data is not matched with the e-commerce scene of country B, and then the unit word "kg" representing weight commonly used in the e-commerce scene of country B may be found as a matching unit word. Further, the unit word of the third type text data china and the number character adjacent to the unit word are modified simultaneously according to the conversion rule of the matching unit word (for example, 2 jin is equal to 1 kg). Specifically, the number character "4" in "4 jin of weight" may be modified to "2", and the unit word "jin" may be modified to "kg", so as to obtain the modified third-type text data, so as to perform the subsequent translation processing.

Of course, if it is determined that the unit word in the third type of text data matches the target electronic market scene, the third type of text data may be used as a target text data without modification. For example, the third type of text data "has a length of 1 meter", wherein the unit word "meter" is also a unit word commonly used in the e-commerce scene of country B, and it can be determined that the unit word in the third type of text data is matched with the e-commerce scene of country B, and no modification processing is performed on the text data.

The above listed unit words of length and weight are only illustrative, and in the specific implementation, the unit words may also include other types of unit words for measurement. Specifically, size unit words (e.g., unit words representing shoe sizes, etc.), currency unit words (e.g., dollars, pounds, etc.), time unit words (e.g., ten days, quarter, etc.), etc. may also be included.

In the present scenario example, it is considered that a data list related to the measurement unit may be further included in the commodity picture in the e-commerce scenario. Such as the size table shown in fig. 7. Wherein, the size table contains third type text data. If the third type of text data after the translation processing is directly backfilled into the size table, the layout effect is often relatively poor, and the sensory experience of consumers is influenced. Therefore, after the server determines the third type of text data, the server may further detect a background pattern around the third type of text data in the target picture, and determine whether a table type background pattern exists around the third type of text data. After determining that the background pattern of the table class exists, the background pattern of the table class around the third type of text data may be recorded, and a corresponding data list tag may be set for the third type of text data. Therefore, when the combined picture is filled back subsequently, the layout of the data list can be adjusted in the picture in a targeted manner according to the background patterns of the peripheral tables of the third type of text data and the data list labels, so that the information content expressed by the data list is clearer, and consumers browsing the data list can obtain better sensory experience.

In the example of the present scenario, it is also considered that the corresponding areas in different e-commerce scenarios are different, and local culture, customs, and abstainment in different areas may be greatly different. For example, it may be preferred to use red as the background color of text data in commodity pictures in country a, while red may be a color considered as unfavorable in the recognition of the B-country textual perception, which is often more objectionable to consumers in country B.

Based on the above situation, when the server specifically processes the OCR recognition result according to the preset processing rule, the server may also determine the culture sensitivity rule matching with the e-market scene in country B. The sensitive rule may be a set of cultural sensitive features including consumer rejections or schemes in an area corresponding to a plurality of different e-commerce scenes, which is established in advance through learning and summarizing local culture, abstaining from and the like in the area. In specific implementation, the culture sensitive rule can be used for carrying out sensitive feature detection on data in the OCR recognition result, and when certain features of the text data in the OCR recognition result are determined to hit the sensitive features in the culture sensitive rule, relevant data in the OCR recognition result can be modified correspondingly in time. Corresponding sensitive prompting labels can be set, so that targeted modification can be carried out during subsequent image combination processing.

For example, when the background color of certain text data in the retrieved OCR recognition result is red, and the red color is found to be a color that is relatively exclusive to consumers in country B according to the culture sensitivity rule, a color-sensitive cue label may be set for the text data, so that the background color of the text data may be adjusted to other colors in time when the drawing process is performed subsequently.

Of course, in specific implementation, the background color parameter of the text data in the OCR recognition result may also be directly modified to another default color, for example, to blue preferred by the consumer in country B. Prompt information can be generated to prompt the merchant A that the background color of the text data triggers the culture sensitive rule, and the background color of the text data is received and modified according to the modification indication of the merchant A.

In this scenario example, it is considered that the picture 1 to be processed is a commodity picture that has been originally applied in the initial e-commerce scenario, i.e., the a country e-commerce scenario. When the merchant a edits and manufactures the picture 1, text data (which may be recorded as fourth type text data) only for the e-commerce scene in country a is set in the picture 1, which may specifically include: the good evaluation of the goods in the shopping website of nation a, the service description promised by the merchant a based on the e-commerce rule of nation a (for example, the rule description about the refund), and the promotion scheme (for example, "buy two-share 8-fold discount" etc.) for the goods displayed on the shopping website of nation a in combination with the marketing strategy of the user, etc. For example, referring to FIG. 7, Picture 1 is directed to consumer-offered "promotions" for the E-commerce scenario of nation A only. The fourth type of text data listed above is often not suitable for being displayed on a shopping website in country B, and a merchant may be required to rewrite new similar text data in combination with the specific situation of country B and the marketing strategy for country B. Therefore, text data of the above type often does not need to be translated and displayed on a commodity picture for e-commerce scenes in country B.

In order to reduce the data processing amount of the subsequent translation processing, in specific implementation, the server may further screen out a fourth type of text data from the OCR recognition result according to a preset filtering rule. Wherein the fourth type of text data comprises at least one of: favorable data based on the initial e-commerce scenario, a service description for the initial e-commerce scenario, a promotional program for the initial e-commerce scenario, and the like.

The preset filtering rule may be a preset rule that includes data features of text data that need to be filtered and are only applicable to the initial e-commerce scene but not applicable to the target e-commerce scene.

In specific implementation, the server may screen out the fourth type of text data by retrieving and comparing data features in OCR recognition results according to the preset filtering rule. Since the fourth type of text data is text data which does not need to be translated, the fourth type of text data can be removed from the OCR recognition result, so that the target text data can be obtained according to the OCR recognition result from which the fourth type of text data is removed.

Furthermore, it is considered that although the fourth type of text data does not need to be translated, the original fourth type of text data in the commodity picture often needs to be deleted during the subsequent image combination processing. Therefore, in specific implementation, the position coordinates of the fourth type of text data can be recorded, so that the original fourth type of text data in the commodity picture can be eliminated in a targeted manner in the subsequent drawing processing.

In the present scenario example, it is considered that data included in the commodity picture in the e-commerce scenario is often complex, for example, the picture may also include an image (e.g., a watermark pattern in the commodity picture) similar to a text character. As a result, when OCR recognition is performed on the picture 1, image data other than the text data may be erroneously recognized as text data (which may be referred to as a fifth type of text data).

Therefore, in order to avoid wasting processing resources and processing time on such text data for subsequent processing, in specific implementation, when the server performs specific processing on the OCR recognition result according to a preset processing rule, the server may determine the confidence of each text data in the OCR recognition result first. The confidence level may be used to characterize a probability value that the recognized text data is true text data, not image data that is erroneously recognized as text data. Further, text data with a confidence level smaller than a preset confidence level threshold value can be screened out from the OCR recognition result to serve as a fifth type of text data. The specific value of the preset confidence threshold can be flexibly set according to specific conditions and precision requirements. Since the fifth type of text data itself is not true text data but image data erroneously recognized as text data, a subsequent translation process is not required. Accordingly, the fifth type of text data may be culled from the OCR recognition results. And acquiring target text data according to the OCR recognition result without the fifth type of text data.

Through the mode, the server carries out corresponding processing on the OCR recognition result by utilizing the preset processing rule matched with the B country electric market scene, and can obtain the target text data matched with the B country electric market scene. It should be understood that the above-mentioned processing method for the OCR recognition result is only an exemplary one. In specific implementation, according to specific conditions, one or more listed processing modes can be flexibly selected according to preset processing rules to process the OCR recognition result so as to obtain target text data matched with the target electronic market. Of course, other types of processing methods besides the above-mentioned processing methods can be introduced to perform corresponding processing on the OCR recognition result.

In this scenario example, after obtaining the target text data matched with the e-market scene in country B, the server may translate the target text data into the target language by using a lexicon (e.g., a common lexicon related to a commodity) accumulated based on the e-market scene in combination with the language characteristics of the e-market scene, so as to obtain more accurate translated target text data.

In specific implementation, the server may call a preset translation model, translate the target text data into a target language, and obtain the translated target text data. The preset translation model comprises a translation model obtained by learning text data in a target e-commerce scene in advance. The preset translation model can be specifically integrated with a word stock accumulated based on the E-market scenes. Therefore, the target text data can be translated into the target language-based accurate target text data which is matched with the electronic market scene and is used as the translated target text data through the preset translation model.

Further, the server can obtain a complete commodity picture which is accurately translated into the target language and is suitable for E-commerce scenes of the B country by combining the translated target text data with the original picture 1 and performing image combination processing.

Specifically, the server may select and perform specific matching processing on the picture 1 and the translated target text data according to a preset matching rule matched with the e-market scene in country B.

The preset drawing combination rule may specifically include a processing rule determined by learning layout and layout of text data and image data in commodity pictures of various different electronic shopping scenes in advance. The layout and typesetting scheme of the text data and the image data in the original commodity picture and the layout and typesetting scheme which are popular in the corresponding E-commerce scene and are more easily accepted by consumers can be synthesized more pertinently based on the preset matching rule to obtain the commodity picture which can be matched with the original picture and can also cause the attention and interest of the consumers in the E-commerce scene of the B country, has relatively good releasing effect and is suitable for the E-commerce scene.

In this scenario example, in a specific implementation, the server may determine, according to a preset drawing combination rule, and according to the position coordinate of the target text data, a picture area where the target text data is located in the target picture. And covering the picture area where the target text data is located by using a background color block with the same color as the background color of the target text data. So as to keep consistent with the background color of the original image. And then laying out the translated target text data above the background color blocks. Thus, the translated target text data can be used to replace the original target text data in the picture 1 by the above-mentioned covering method instead of the erasing method. But actually, the original target text data is still reserved below the background color block of the image, so that the subsequent merchant A can conveniently acquire the original target text data again according to specific needs.

In the scene example, when the server lays out the translated target text data according to the preset drawing combination rule, the character characteristics of the translated target text data can be determined according to the character characteristics of the target text data; and laying the translated target text data according to the character characteristics of the translated target text data. For example, the translated target text data is laid out in the same character size according to the character size of the target text data characters. Or, the translated target text data is laid out in the same color according to the character color in the target text data, and so on.

In the scene example, when the server lays out the translated target text data according to the preset drawing combination rule, the size of the region of the picture region where the target text data is located in the target picture and the number of characters contained in the translated target text data can be determined; and adjusting character characteristics of the translated target text data according to the area size of the picture area where the target text data is located and the number of characters contained in the translated target text data. For example, the size of the region of the picture in which the target text data is located is 10 square centimeters, the original target text data contains 10 characters, and the character size of each character is number 4. But the number of characters contained in the translated target text data becomes 20. At this time, the server can reduce the character size of the characters of the translated target text data into 10 numbers according to a preset drawing combination rule, and then lay the translated target text data in the same picture area of 10 square centimeters according to the reduced characters, so that the layout of the translated target text data in the picture is relatively more reasonable and attractive, and consumers can obtain better sensory experience.

In this scenario example, in a specific implementation, the server may further detect whether the translated target text data carries a data list tag. If the data list tag carried by the target text data is determined, the text data can be determined to be the third type of text data in the data list. Further, the character characteristics of the text data can be adjusted in a targeted manner according to the background patterns of the peripheral tables of the text data, so that the data list displayed in the commodity picture after the drawing processing is relatively clearer and more beautiful, and the sensory experience of a consumer during reading is further improved. Referring to fig. 8, the size table in the figure is a size table suitable for the B country e-commerce scenario obtained after the adjustment processing in the above manner.

In this scenario example, the size table in fig. 8 is relative to the size table in picture 1, and the unit words in the table have been processed and translated into matching unit words applicable to country B. And, according to the conversion rule of the matching unit word, the corresponding conversion processing is also performed on the digital character adjacent to the unit word (for example, the digital character before the unit word), and the data character is modified correspondingly according to the conversion processing result.

Specifically, for example, since the unit words used in countries a and B for representing Height (Height) are the same unit words representing the same amount, the corresponding conversion rule is: 1 cm-1 cm. Therefore, the numeric characters preceding the unit word in the third type of text data for representing the height in the size table of fig. 8 are not modified with respect to the size table of fig. 7, and the original numeric characters are used; the unit word is translated from cm into matching unit word cm suitable for B country.

For another example, the unit words representing the Weight (Weight) used in nation a and nation B are unit words representing different quantities. Specifically, the unit word used in country a for weight is "jin", which corresponds to a weight of 500 g, and the unit word used in country B for weight is "kg" (kg), which corresponds to a weight of 1000 g. It can be seen that the above two unit words are unit words that represent different quantities. Therefore, when the third type of text data for representing the body weight in the size table is specifically processed, the unit word is modified from 'jin' to a matching unit word 'kg' suitable for the country B; meanwhile, the conversion rule of the matched unit word (for example, 1 kilogram to 0.5 kilogram) is obtained and is used for carrying out conversion processing on the digital characters before the unit word; and then according to the conversion processing result, correspondingly modifying the digital characters before the unit word. Specifically, for the third type of text data of "90 jin or less", except that the unit word "jin" is modified to match the unit word "kg"; and converting the digital character 90 before the unit word according to the conversion rule of the matched unit word to obtain a corresponding conversion processing result 45, modifying the original digital character 90 into 45 according to the conversion processing result, and obtaining the processed third-type text data below 45 kilograms. And then, the third type of text data after the above processing is translated, and finally, translated text data "Under 45 kg" shown in fig. 8 is obtained. The obtained size table is accurate and is suitable for E-commerce scenes of the B country.

In this scenario example, when laying out the translated target text data according to a preset drawing rule, the server may further detect whether the translated target text data is provided with a vertical arrangement tag. In order to achieve the effect that the original image is desired to be highlighted and to match the e-commerce scenario of country B and conform to the reading habit of the consumer of country B under the condition that it is determined that the translated target text data is provided with the vertically arranged tags, see fig. 8. The translated target text data may be rotated by 90 degrees clockwise to obtain the rotated target text data. And then laying the rotated target text data above the background color block at the corresponding position.

In the scene example, when the server lays out the translated target text data according to the preset drawing combination rule, whether data carries a sensitive prompt tag in the OCR recognition result can also be detected. When it is determined that the data carrying the sensitive prompt tags exists in the OCR recognition result, the corresponding modification may be performed on the partial data in combination with culture, customs, abstinence and the like of country B in the process of combining the drawing so as to better meet the e-commerce scene of country B.

For example, when it is detected that the partially translated target text data carries a color-sensitive prompt tag for a background color, the original background color of the target text data can be automatically adjusted to other colors which are more acceptable to the national culture B according to a preset drawing matching rule, so that the processed commodity picture can be effectively prevented from containing elements which are rejected by the national culture B, and the putting effect of the processed commodity picture on the national culture B is further improved.

Through the mode, the server can perform related drawing processing such as layout, splicing, combination and the like on the translated target text data and the picture 1 by using a preset drawing rule matched with the E-commerce scene of the country B to obtain a commodity picture which accurately translates the target text data into English and is suitable for the E-commerce scene of the country B. It should be understood that the above-listed combination is only a schematic illustration. In specific implementation, according to specific preset drawing rules, one or more of the listed drawing modes can be flexibly selected to be combined for drawing processing according to specific conditions, so as to obtain a final commodity picture after drawing processing. Of course, other types of matching manners besides the listed matching manner may be introduced to perform more targeted matching processing on the translated target text data and the picture 1.

After the server completes the image combination processing according to the above, and obtains the corresponding commodity image (which may be recorded as image 2), the image 2 may be sent to the terminal device in a wired or wireless manner. And the terminal equipment can show the picture 2 to the business A. Subsequently, the merchant A can directly launch the picture 2 obtained after the processing to the commodity detail page of the shopping website of the B country so as to introduce and promote the No. 1 commodity on the shopping website of the B country.

Therefore, although the merchant A does not know English, does not know specific E-commerce scenes of the country B and local culture, custom and other characteristics of the country B, the target text data to be translated in the picture 1 can be accurately translated into the target language by the mode, meanwhile, the E-commerce scenes suitable for the country B are obtained, the good putting effect is achieved, and the commodity pictures published on the shopping website of the country B can be directly put in.

In another specific example of the scene, the translated target text data in picture 2, and the image data in picture 2 are presented in picture 2 in a mutually independent manner. Further, the translated target text data and image data in the picture 2 also support operations such as modification of merchants. Therefore, if the merchant A has certain understanding on the E-commerce scene of the country B, the merchant A can further modify and optimize on the basis of the picture 2 by combining the factors of marketing strategy, aesthetic feeling and the like of the merchant A, so that the commodity pictures subsequently put on the shopping website of the country B better meet the demands of the merchant A and obtain better putting effect.

Specifically, as shown in fig. 9, when receiving the picture 2 sent by the server, the terminal device displays the picture 2 to the merchant a, and also responds to the operation of the merchant a and displays a picture modification request interface to the merchant a. Since the translated target text data and the image data in picture 2 are relatively independent and edit-enabled, they are shown in picture 2. Therefore, the terminal device can provide an interface for modifying and editing the translated analog text data in the picture 2 and the image data for the business A through the picture modification request interface.

As shown in fig. 9, the image modification request interface may specifically include: and the operation areas comprise a picture preview area, a layer area, a preview area, an editing area and the like, and are respectively associated with the translated target text data and the image data in the picture 2. The merchant A can conveniently modify and edit the translated text data in the picture 2 and the image data respectively by performing corresponding operations in the operation area.

Referring to fig. 9, the picture preview area is located at the leftmost side of the picture modification request interface, and through the operation area, the business nail can conveniently preview a picture to be edited, such as picture 2. The pictures to be edited can be switched by mouse clicking or other operations in the operation area.

The layer area is located on the right side of the picture preview area, and the picture 2 in a modifiable and editing state can be displayed to the business A specifically through the operation area. For example, the translated text data in picture 2, and the image data in the operation area are displayed in a relatively independent and edit-supporting manner.

The preview area is located in the middle of the picture modification request interface, the picture 2 effect can be displayed to the first merchant through the operation area, and the original commodity picture before processing, namely the picture 1, can be switched through the operation area. Therefore, the first merchant can compare the two pictures and modify the picture 2 in a targeted manner based on the comparison result.

The editing area may include an area located at the right side and the top of the picture modification request interface and used for performing a specific editing operation. The user can perform specific operation in the operation area according to the corresponding prompt.

In this scenario example, the business a may perform one or more modification operations listed below on the picture 2 through the above-mentioned picture modification request interface. For example, a modification operation on the content of the text data may be performed on the translated target text data in the picture 2 (which may also include text data which is removed from the OCR recognition result and is not translated), and/or a modification operation on character features in the text data. For example, the font, font size, color, alignment, bolding, italics, position, etc. of characters in the text data may be modified. Operations such as modifying the color of the background color block, modifying the size of the background color block, adjusting the position of the background color block, and the like can also be performed on the background of the text data in the picture 2. The operations of adding text data and adding background color blocks can also be performed on the basis of the existing translated text data and image data of the picture 2. And the translated text data and/or the background color blocks can be quickly adjusted to other positions in the picture 2 in a copying or pasting mode. The modified picture 2 can also be downloaded and saved to the local of the terminal device. The picture scale of the current preview zone can also be changed by a zoom operation to better check picture 2. And a reduction operation can be carried out to reduce the picture 2 into the unprocessed picture 1, so that the business A processes the picture 1 again in the convenient area according to own ideas, and the like.

Specifically, the merchant a may supplement, by adding text data and the like, the service description provided by the merchant a based on the e-commerce rule of country B on the basis of the picture 2, a new sales promotion scheme (for example, "top 50 buyers share half-price advantage" and the like) for the consumer in the e-commerce scenario of country B, and the like, according to the marketing strategy of the store for the e-commerce scenario of country B by using the above-mentioned picture modification request interface.

Of course, the modification operations listed above are only illustrative. In specific implementation, according to specific conditions, the merchant A can also perform other types of modification operations through the displayed picture modification request interface according to the needs of the merchant A. The present specification is not limited to these.

The terminal device may receive a modification operation initiated by the merchant a through the image modification request interface, and generate a modification instruction for the translated target text data in the image 2 and/or the image data. And then, according to the modification instruction, the translated target text data in the picture 2 and/or the image data can be modified correspondingly.

Therefore, the first merchant can flexibly modify and adjust the translated target text data in the picture 2 and/or the image data according to own ideas on the basis of the processed picture 2 fed back by the server to edit and optimize the picture 2, so that the finally obtained picture 2 better meets the requirements of the first merchant, is more suitable for E-commerce scenes of the country B to be released, and can meet the personalized processing requirements of the first merchant.

In another specific scenario example, in order to further facilitate the merchant a to release a commodity picture for the commodity No. 1, after obtaining the processed picture 2 fed back by the server, the terminal device may further respond to the operation of the merchant a to display a picture release interface to the merchant a. The terminal equipment can be connected with a server of a B country shopping website, and the picture putting and publishing interface is associated with a page of the B country shopping website. Specifically, the image release interface may include a custom address input box and a release confirmation key.

As shown in fig. 2, the server responsible for processing the picture 1 may be directly connected to the server of the B country shopping website, so that the server may receive and respond to the issuing instruction of the merchant a, and more efficiently issue the processed picture 2 suitable for the B country e-commerce scene to the B country shopping website, thereby simplifying the issuing operation of the merchant and improving the issuing efficiency.

In specific implementation, the merchant a can input the picture address (for example, url address) of the picture 2 in the detail page of the B country shopping website into the self-defined address input box, and click the release confirmation key to complete the operation. And the terminal equipment can receive and respond to the operation, generate and send a corresponding issuing instruction to a server of the B country shopping website, wherein the issuing instruction can specifically carry the picture address and the picture 2 input by the A. The server of the B country shopping website responds to the issuing instruction, can find the picture position corresponding to the picture address in the detail page of the No. 1 commodity of the B country shopping website according to the picture address, and call the corresponding interface, replace the original picture 1 on the picture position with the picture 2, thus can simplify the operation of the merchant A, finish the issuing of the commodity picture about the No. 1 commodity for the merchant A automatically and conveniently with high efficiency.

Of course, in specific implementation, the picture 1 itself may also carry address data associated with the picture position of the B country shopping website, or an identity (e.g., a picture number on the B country shopping website, etc.) that can correspond to the picture position. Therefore, when the merchant A determines to publish, the operation can be completed only by triggering the publication confirmation key at the midpoint of the picture release and publication interface. Correspondingly, the terminal device may generate and send a corresponding issuing instruction, where the issuing instruction carries the picture 2, and the address data or the identity obtained from the picture 1. After receiving the issuing instruction, the server of the B country shopping website can accurately find the corresponding picture position in the B country shopping website according to the address data or the identity, and further replace the original picture at the picture position with the picture 2. Therefore, the operation of the merchant A can be further simplified, and the processed picture 2 can be more efficiently and conveniently released to a shopping website of the country B.

Referring to fig. 10, an embodiment of the present disclosure provides a method for processing a commodity picture. In particular implementations, the method may include the following.

S1001: and acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to the commodity and text data related to the commodity.

In some embodiments, the processing method of the commodity picture may be specifically applied to a server, and may also be applied to a terminal device with data processing capability. In the present embodiment, the application to the server is specifically described as an example. For the case of application to a terminal device, reference may be made to an embodiment applied to a server. Therefore, the description is not repeated.

In some embodiments, the target picture may specifically include a commodity picture to be processed, which is suitable for the initial e-commerce scene. For example, the target picture may be a picture for introducing and promoting the related goods in the goods detail page of the initial e-commerce scene. The target picture may specifically include image data related to the commodity and text data related to the commodity. The initial electronic market scene is specifically understood as a shopping website or a shopping APP and the like which are originally correspondingly suitable for the target picture.

Specifically, the image data may include: the image of the commodity body, the scene image when the commodity is used, the promotion image related to the commodity and the like. The text data may specifically include: text data for describing commodity parameter information, text data for describing commodity use conditions, text data for popularizing commodities, text data introducing each part of a commodity body and the like. Of course, the image data and the text data listed above are only an illustrative description. In specific implementation, the image data and the text data may also include other types of image data and text data related to the product according to the specific situation of the product. The present specification is not limited to these.

In some embodiments, the target picture may be a commodity picture suitable for an initial e-commerce scenario. The language used in the target picture is a language (which can be marked as an initial language) common to the initial e-commerce scene, and the layout, the typesetting and the like of the text data, the image data, the text data and the image data in the picture are also suitable for the initial e-commerce scene and are matched with the local culture of the area corresponding to the initial e-commerce scene. Therefore, the target picture can often generate better attraction for the consumers in the area corresponding to the initial e-commerce scene, and has better delivery effect.

In this embodiment, the target picture needs to be processed correspondingly so as to be translated into a commodity picture which uses the target language and is suitable for the target e-commerce scene. Therefore, the processed commodity pictures can be released to the target E-commerce scene in the following process, related commodities are introduced and popularized to the consumption of the target E-commerce scene, and the commodity list rate in the target E-commerce scene is improved.

The target language may specifically include a language different from the initial language. The target e-market scene may specifically include a shopping website or a shopping App using a target language as a general language.

For example, the initial e-commerce scenario is a shopping APP in nation a, and the initial language may be the language used on the shopping APP in nation a. The target e-market scene may be a B country shopping website and the target language may be a B country language used on the B country shopping website.

In some embodiments, the parameter indicating the target language may specifically include a parameter indicating a language of the target language. Such as english, chinese, french, etc.

In some embodiments, the merchant who wants to sell the merchandise in the target e-market scene may use the existing merchandise picture used in the initial e-market scene before the merchant's store to introduce and promote the merchandise as the target picture. And at least inputs or sets the indication parameters of the target language used by the targeted e-commerce scenario. And then can generate and send corresponding picture processing requests to the server. The picture processing request may carry the target picture and an indication parameter of the target language.

Certainly, when the image processing request is specifically generated, the merchant may also set a specific target e-market scene, for example, a TB shopping network in country B; it is also possible to set a treatment style desired by oneself, such as "lovely style", "gorgeous style", and the like. And further more detailed picture processing requests can be generated and sent to the server.

The server receives the picture processing request, and can acquire and wait for the target picture and the target language indication parameter through data analysis. Further, the server can determine the target language for translating the target picture according to the target language indication parameters.

S1002: and presetting the target picture to acquire target text data matched with a target electronic scene using a target language from the target picture.

In some embodiments, the server may automatically determine, according to the target language, a target e-commerce scene to which the e-commerce scene in the target language is used as the processed commodity picture.

In some embodiments, the preset processing may specifically include data processing performed on the target picture before the text data is subjected to the translation processing. The preset processing is matched with the target electric market scene, and target text data matched with the target electric market scene can be acquired from the target picture by performing the preset processing on the target picture.

In some embodiments, the target text data may specifically include text data that needs to be translated in the target picture of the target electronic market.

In some embodiments, the preset processing on the target picture to obtain the target text data matched with the target electronic market using the target language from the target picture may include the following steps: and carrying out image text recognition on the target picture to obtain an image text recognition result. And correspondingly processing the image text recognition result according to preset processing rules to obtain the target text data, wherein the preset processing rules comprise processing rules matched with the target E-market scene.

In some embodiments, the image text recognition may specifically include OCR recognition (optical character recognition) and the like. Of course, the above-listed OCR recognition is only an illustrative illustration. In specific implementation, according to specific situations, other suitable manners capable of recognizing and extracting text data from a picture may be adopted to perform image text recognition. The present specification is not limited to these.

In some embodiments, when implemented, text data included in the target picture may be identified and extracted by performing image text recognition on the target picture, and other data related to the text data, such as position coordinates of the text data in the target picture, a background color of the text data in the target picture, character features of the text data (for example, an arrangement direction of characters, a size of the characters, a color of the characters, and the like).

Correspondingly, the image text recognition result obtained by the image text recognition may specifically be one or more of the data: text data in the target picture, position coordinates of the text data in the target picture, a background color of the text data in the target picture, character features of the text data, and the like. Of course, the image text recognition results listed above are only illustrative. In specific implementation, the image text recognition result may further include other types of data obtained through image text recognition and other processing according to specific situations and processing requirements. For example, image features of image data adjacent to the text data in the target picture, and the like may also be included.

In some embodiments, the preset processing rule may specifically include a processing rule matching the target e-market scene. Before specific implementation, the commodity pictures in shopping websites or shopping platforms such as shopping APP (application) corresponding to different e-commerce scenes can be learned in advance, and processing rules respectively corresponding to the different e-commerce scenes are established and obtained by learning local culture, language, custom and the like of areas corresponding to the different e-commerce scenes. Wherein each processing rule can be respectively associated with one or more e-market scene objects. The processing rule can be used for integrating two dimensions of business and culture related to the corresponding electronic market scene to perform targeted processing on the image text recognition result so as to obtain target text data matched with the corresponding electronic market scene. And then, a commodity picture which is suitable for the e-commerce scene and is easily accepted by consumers in the e-commerce scene and has a better delivery effect in the e-commerce scene can be obtained based on the target text data.

In some embodiments, the performing, according to a preset processing rule, the corresponding processing on the image text recognition result may include: screening out text data which do not need to be translated from the image text recognition result according to a preset processing rule, and removing the text data; and/or determining text data needing to be translated from the image text recognition result, and reserving the text data so as to translate the text data into a target language subsequently. The corresponding processing performed on the image text recognition result may be specifically one processing, or may be a combination of multiple types of processing. By correspondingly processing the image text recognition result, target text data which needs to be translated subsequently and is matched with the target electronic market scene can be acquired from the image text recognition result.

In some embodiments, further, the specific features of the target text data may be detected and determined, and corresponding data tags may be set according to preset processing rules, so that during subsequent image combination processing, some target text data may be subjected to targeted region processing according to the data tags, and thus, a commodity image with a relatively better delivery effect may be obtained.

In some embodiments, the above correspondingly processing the image text recognition result according to a preset processing rule may include the following steps: determining text characteristics of text data in an image text recognition result; screening out text data with text characteristics conforming to first-class text characteristics from the image text recognition result as first-class text data; wherein the first type of text data comprises: the text data is arranged on the commodity body, and/or the text data is arranged on the trademark of the commodity; and removing the first type of text data from the image text recognition result.

In some embodiments, the first type of text data may specifically include text data set on a product body, for example, text data printed on a product package, or text data included in a label or a hang tag of a product, and may further include text data on a trademark of a product, and the like. For such text data, most consumers want to be what they see, i.e., what the text data are seen in the commodity pictures on the shopping website or shopping APP platform, etc., is the same as the text data of the commodity entity they have taken, and belongs to the text data that does not need to be translated. Therefore, it can be determined that the first type of text data does not belong to the target text data, and the first type of text data is removed from the image text recognition result, so that the target text data obtained based on the image text recognition result does not contain the first type of text data.

In some embodiments, image features associated with the text data may be determined first according to the image text recognition result. For example, a background pattern of text data in a picture of a commodity, and the like. And then screening out the first type of text data from the image text recognition result according to the image characteristics related to the text data. For example, text data with the background pattern in the commodity picture as the commodity body pattern can be screened from the image text recognition result according to the background pattern of the text data in the commodity picture as the first type of text data. In addition, text data with a background pattern having a high similarity to the trademark pattern of the commodity may be selected from the image text recognition result based on the background pattern of the text data in the commodity picture as the first type text data. Of course, the manner in which the first type of text data is determined as listed above is merely illustrative. In specific implementation, according to specific situations, the first type of text data may be screened from the image text recognition result by using other suitable methods. The present specification is not limited to these.

In some embodiments, the above correspondingly processing the image text recognition result according to a preset processing rule may further include the following steps in specific implementation: determining multi-line text data belonging to the same text region according to the position coordinates of the text data in the image text recognition result; plural lines of text data belonging to the same text area are merged into one text data.

In some embodiments, in order to make the subsequent translation more accurate and facilitate layout and layout in the subsequent image combination process, adjacent lines of text data representing the same piece of content may be screened from the image text recognition result according to the position coordinates of each text data. For example, according to the position coordinates of the text data in the image text recognition result, multiple lines of text data with adjacent position coordinates, same character characteristics and same length are found as multiple lines of text data to be combined. And then, through multi-line combination, splicing and combining the multi-line text data to obtain a total text data, so that the total text data can be translated integrally during subsequent translation processing, and the translation is more accurate and reliable. And subsequently splitting the translated total text data into a plurality of lines of text data with the same length for layout.

In some embodiments, the above correspondingly processing the image text recognition result according to a preset processing rule may further include the following steps in specific implementation: screening out text data with vertically arranged characters from the image text recognition result as second type text data according to character features of the text data; and converting the second type text data into text data arranged along a preset character arrangement direction (for example, a conventional character arrangement direction: horizontal arrangement) according to the character arrangement direction in the second type text data, wherein the text data is used as the processed second type text data, and a vertical arrangement label is arranged in the processed second type text data. Therefore, the second type text data which is arranged along the preset character arrangement direction, is relatively standard and is suitable for being translated after being processed can be obtained subsequently. And then, translation processing is carried out based on the processed second-class text data, so that errors in translation are reduced. Meanwhile, during subsequent image combination processing, the characters in the translated second type text data can be vertically arranged in a targeted manner according to the vertical arrangement labels arranged in the processed second type text data, so that the original layout form is restored, and the original putting effect is achieved.

Of course, in a similar manner, text data with other character characteristics (e.g., the color of the character, the size of the character, etc.) different from the conventional character characteristics can also be found as the second type of text data in specific implementation. And processing the second type of text data to obtain more standard processed second type of text data suitable for translation, and setting a data tag corresponding to character features of characters contained in the text data.

The second type of text data may be specifically understood as text data in which character features in the commodity picture are different from conventional preset character features. Specifically, the text data in which the character features of the included characters are vertically arranged, or the text data in which the character features of the included characters are radially arranged, or the text data in which the character features of the included characters are larger than a preset threshold value, may be included. The second type of text data is also a text data which needs to be translated, but contains characters with special character characteristics. Therefore, the screened processed second type text data can be retained as a kind of target text data.

In some embodiments, the above correspondingly processing the image text recognition result according to a preset processing rule may include the following steps: screening out text data containing unit words from the image text recognition result through unit word retrieval to serve as third type text data; determining whether the unit words in the third type of text data are matched with a target E-market scene; under the condition that the unit words in the third type of text data are determined to be not matched with the target E-market scenery, determining matched unit words corresponding to the unit words; and modifying the unit words in the third type text data and the numerical characters adjacent to the unit words according to the conversion rule of the matched unit words to obtain the processed third type text data.

The third type of text data may specifically include text data including unit words related to metering descriptions related to the commodities. The third type of text data is also a text data that needs to be translated, but contains unit words that are not suitable for the target electronic landscape.

In this embodiment, the unit words used in the third type of text data are applicable to the initial e-commerce scenario, but not necessarily applicable to the target e-commerce scenario. If the third type of text data is directly translated without modification, and the third type of text data is backfilled and combined into a commodity picture and displayed to a consumer of a target e-commerce scene, the consumer is often puzzled, and the release effect is influenced.

Therefore, in specific implementation, it may be determined whether the unit word included in the third type of text data matches the target e-market scene, for example, whether the unit word is a unit word commonly used by the target e-market scene, according to a preset processing rule. Under the condition that the unit words are determined to be not matched with the target E-commerce scene, the unit words suitable for the target E-commerce scene can be found and used as matched unit words. And modifying the unit words used in the original text in the third type of text data by using the matched unit words. For example, the unit word "kilogram" originally used in the third type of text data is modified into the matching unit word "kilogram" suitable for the target e-commerce scene.

Further, it is considered that the matching unit words and the originally used unit words may have different expressed quantities. For example, the weight units of "jin" and "kg" are not equal, and the monetary values of the currency units of "U.S. dollars" and "pounds" are not equal. Therefore, the conversion process may be performed according to the conversion rule of the matching unit word (for example, the conversion ratio between the matching unit word and the unit word originally used), so that the unit word in the third-type text data is modified into the matching unit word, and the digital character adjacent to the unit word (for example, the digital character adjacent to the unit word and located in front of the unit word) is subjected to the corresponding modification process according to the conversion process result, thereby obtaining the processed third-type text data. And then the modified third-class text data is used as target text data to perform subsequent translation processing.

In contrast, when it is determined that the unit words included in the third type of text data match the target electronic market, the third type of text data may be directly used as a target text data without modifying the third type of text data.

In some embodiments, it is considered that for the pictures in the e-commerce scene such as the commodity pictures, a data list related to the measurement units may be included. Such as a size scale, weight scale, etc. Such data lists often contain a plurality of text data of the third type. In order to better process such data lists, in a specific implementation, after the third type of text data is determined, further, a detection may be performed on a background pattern around the third type of text data in the target picture, so as to determine whether a table-type background pattern exists around the third type of text data. After determining that the background pattern of the table class exists, the background pattern of the table class around the third type of text data may be recorded, and a corresponding data list tag may be set for the third type of text data. When the combined picture is filled back subsequently, the layout of the data list can be adjusted in the picture in a targeted manner according to the background patterns of the tables and the data list labels around the third type of text data (for example, the size of the characters of the third type of text data arranged in the tables or the size of the background patterns of the tables is modified and the like), so that the information content expressed by the data list is clearer, and a consumer browsing the data table can obtain better sensory experience.

In some embodiments, the above correspondingly processing the image text recognition result according to a preset processing rule may further include the following steps in specific implementation: determining a target culture sensitive rule matched with a target E-market scene; and correspondingly modifying the data in the image text recognition result according to the target culture sensitivity rule.

The cultural sensitivity rule may be a set of sensitivity features (for example, features such as red is used as a background color, or the number 13 is used, or there is no interest in cattle) which are established in advance by learning and summarizing local culture, abstinence and the like of areas corresponding to a plurality of different e-commerce scenes and include consumer rejections or solutions of the areas. Typically a culture sensitive rule may correspond to one or more e-commerce scenarios. The target culture sensitive rule is a culture sensitive rule corresponding to the target E-market scene.

In specific implementation, the target culture sensitive rule can be used for detecting the sensitive features of the data in the image text recognition result, and when some features of the text data in the image text recognition result are determined to hit the sensitive features in the target culture sensitive rule, the relevant data in the image text recognition result can be modified correspondingly in time. Or, the relevant data may be provided with a corresponding sensitive prompt tag without modification, so that the subsequent image combination process may be modified in a targeted manner.

In some embodiments, the above correspondingly processing the image text recognition result according to a preset processing rule may further include the following steps in specific implementation: according to a preset filtering rule, fourth type text data are screened out from the image text recognition result, wherein the fourth type text data comprise at least one of the following text data: the method comprises the following steps of (1) based on favorable data of an initial e-commerce scene, service description aiming at the initial e-commerce scene, and a promotion scheme aiming at the initial e-commerce scene; and removing the fourth type of text data from the image text recognition result. The preset filtering rule may be a preset rule that includes data features of text data that need to be filtered and are only applicable to the initial e-commerce scene but not applicable to the target e-commerce scene.

The fourth type of text data may be specifically only suitable for text data of which the initial e-commerce scene is not suitable for the target e-commerce scene. For example, the favorable data obtained in the initial e-commerce scene is either a service description for the initial e-commerce scene (e.g., a description about "three-pack" service, etc.), or a promotion scheme for the initial e-commerce scene (e.g., "full two-fold out of eight" coupon information, etc.), etc. The fourth type of text data does not need subsequent translation processing and does not belong to target text data. Moreover, the fourth type of text data is different from the first text data, and does not need to appear in commodity pictures put into a target e-commerce scene in the future. Therefore, in specific implementation, the fourth type of text data may be removed from the image text recognition result, so that more accurate target text data may be obtained based on the image text recognition result from which the fourth type of text data is removed.

In some embodiments, the above correspondingly processing the image text recognition result according to a preset processing rule may further include the following steps in specific implementation: determining the confidence coefficient of the text data in the image text recognition result; screening out text data with the confidence coefficient smaller than a preset confidence coefficient threshold value from the image text recognition result to serve as fifth type text data; and removing the fifth type of text data from the image text recognition result. The confidence level may be used to characterize a probability value that the recognized text data is true text data, not image data that is erroneously recognized as text data. The specific value of the preset confidence threshold can be flexibly set according to specific conditions and precision requirements.

The fifth type of text data body may specifically include image data that is erroneously recognized as text data in an image text recognition process. For example, a graphic or pattern in the target picture that approximates the text data (e.g., a watermark in the target picture, etc.). The fifth type of text data does not need to be translated, and does not belong to the target text data. Therefore, in specific implementation, the fifth type of text data may be removed from the image text recognition result, so that more accurate target text data may be obtained based on the image text recognition result from which the fifth type of text data is removed.

By the method, the image text recognition result can be correspondingly processed according to the preset processing rule matched with the target electric market scene, so that target text data which needs to be translated and is matched with the target electric market scene is obtained. It should be understood that the above-mentioned processing method for the image text recognition result is only an exemplary description. In specific implementation, according to specific conditions, one or more listed processing modes can be flexibly selected according to preset processing rules to process the image text recognition result so as to obtain target text data matched with the target electronic market. Of course, other types of processing methods besides the above-mentioned processing methods can be introduced to perform corresponding processing on the image text recognition result.

S1003: and translating the target text data into a target language to obtain translated target text data.

In some embodiments, in specific implementation, only the target text data may be translated, and the target text data is translated into the target language to obtain the translated target text data. This may result in translated target text data suitable for the target electronic marketplace view.

In some embodiments, the translating the target text data into the target language to obtain the translated target text data may include the following steps: and calling a preset translation model, translating the target text data into a target language, and obtaining the translated target text data, wherein the preset translation model comprises a translation model obtained by learning the text data in the target e-commerce scene in advance.

Before specific implementation, text data related to the last commodity in the target e-commerce scene and based on the target language can be obtained in advance to serve as training data, and then a preset translation model for the target e-commerce scene can be obtained through learning and training the training data. Therefore, the preset translation model can be close to the E-commerce prospect, and translation processing can be performed on target text data more accurately.

In some embodiments, a thesaurus for e-commerce scenarios may also be pre-established before implementation. And when the target text data is translated, the word stock aiming at the E-market scene can be utilized to translate the target text data more accurately.

S1004: and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

In some embodiments, the combining process is performed according to the translated target text data and the target picture to obtain a commodity picture suitable for the target electronic market scene, and the specific implementation may include the following steps: and carrying out image combination processing on the target picture and the translated target text data according to preset image combination rules to obtain the commodity picture suitable for the target E-commerce scene, wherein the preset image combination rules comprise typesetting rules matched with the target E-commerce scene.

The preset drawing combination rule may specifically include a processing rule determined by learning layout and layout of text data and image data in a commodity picture of a target e-commerce scene in advance. The layout and typesetting scheme of the text data and the image data in the original target picture and the layout and typesetting scheme which are popular in the target e-commerce scene and are more easily accepted by consumers can be synthesized pertinently based on the preset matching rule, so that the commodity picture which can be matched with the original target picture, can arouse the attention and interest of the consumers in the target e-commerce scene, has relatively good putting effect and is suitable for the target e-commerce scene is obtained.

In some embodiments, the combining processing on the target picture and the translated target text data according to a preset combining rule may include the following steps: determining a picture area where the target text data in the target picture is located according to the position coordinates of the target text data; covering a picture area where the target text data is located by using a background color block with the same color as the background color of the target text data; and laying the translated target text data above the background color blocks. Thus, the translated target text data can be used to replace the original target text data in the target picture in an overlaying mode rather than a clearing mode. However, the original target text data is actually reserved under the background color block of the image, so that the original target text data can be conveniently acquired again if needed subsequently.

In some embodiments, when the translated target text data is laid out according to a preset drawing combination rule, the character features of the translated target text data can be determined according to the character features of the target text data; and laying the translated target text data according to the character characteristics of the target text data. For example, the translated target text data may be laid out in the same character size according to the character size of the character that the target text data originally contained in the target drawing. Alternatively, the translated target text data may be laid out in the same color according to the character color in the target text data, and the like.

In some embodiments, when the translated target text data is laid out according to a preset drawing combination rule, the region size of a picture region where the target text data is located in the target picture and the number of characters contained in the translated target text data can be determined; and adjusting character characteristics of the translated target text data according to the area size of the picture area where the target text data is located and the number of characters contained in the translated target text data.

For example, it may be determined that the area size of the picture area where the target text data is located is 10 square centimeters, the target text data in the original target picture includes 10 characters, and the character size of each character is number 4. But the number of characters contained in the translated target text data becomes 20. At this time, the character size of the character of the translated target text data can be adaptively reduced to 10 according to the preset drawing combination rule, and the translated target text data is laid in the same picture area of 10 square centimeters according to the reduced character, so that the layout of the translated target text data in the obtained commodity picture is relatively more reasonable and attractive, and a better sensory experience can be obtained when a consumer reads the commodity picture.

In some embodiments, when the graph combination is performed according to a preset graph combination rule, it may be further detected whether the translated target text data carries a data list tag. If the data list tag carried by the target text data is determined, the text data can be determined to be the third type of text data in the data list. Further, the character characteristics of the text data can be adjusted in a targeted manner according to the background patterns of the peripheral tables of the text data, so that the data list displayed in the commodity picture after the drawing processing is relatively clearer and more beautiful, and the sensory experience of a consumer during reading is further improved. As can be seen in fig. 8.

In some embodiments, when laying out the translated target text data according to a preset drawing combination rule, it may also be detected whether the translated target text data is provided with a vertically arranged tag. Under the condition that the translated target text data is determined to be provided with the vertically arranged tags, rotating the translated target text data by 90 degrees in the clockwise direction to obtain the rotated target text data; and laying the rotated target text data on the target picture. As shown in fig. 8. Therefore, the effect that the original target picture is required to be highlighted can be achieved, and meanwhile, in order to better match the target E-commerce scene, the reading habit of the consumer of the target E-commerce scene is matched, and a better putting effect is obtained.

In some embodiments, when the translated target text data is laid according to a preset drawing combination rule, whether the image text recognition result has data carrying a sensitive prompt tag or not can be detected. When it is determined that the data carrying the sensitive prompt tag exists in the image text recognition result, corresponding modification and adjustment can be performed on the part of data in combination with culture, customs, abstinence and the like of a local area corresponding to the target e-commerce scene in the graph combining process, so that the modified and adjusted commodity picture is more easily approved and accepted by consumers of the target e-commerce scene.

By the mode, the translated target text data and the target picture can be subjected to related picture combining processing by utilizing the preset picture combining rule matched with the target electronic market scene, so that the commodity picture which is capable of accurately translating the target text data into the target language and is also well suitable for the target electronic market scene is obtained. It should be understood that the above-listed combination is only a schematic illustration. In specific implementation, according to a preset drawing combination rule, one or more listed drawing combination modes can be flexibly selected to perform drawing combination processing according to specific conditions, so as to obtain a commodity picture after the drawing combination processing. Of course, in specific implementation, other types of matching manners besides the listed matching manners may be introduced to perform more targeted and more detailed matching processing on the translated target text data and the target picture.

In the embodiment of the specification, the target picture to be translated into the target language is subjected to preset processing corresponding to the target E-market scene, so that target text data matched with the target E-market scene is obtained from the target picture; then, the target text data is translated into a target language in a targeted manner to obtain translated target text data; and carrying out image combination processing according to the translated target text data and the target picture to obtain a commodity picture suitable for the target electronic market scene. Therefore, the text data in the target picture can be accurately translated into the target language, and meanwhile, the commodity picture which is suitable for the target e-commerce scene and has a good putting effect is obtained.

In some embodiments, the image text recognition result is correspondingly processed according to a preset processing rule to obtain the target text data, and the specific implementation may include the following contents: determining and removing text data which does not need to be translated into a target language from the image text recognition result according to a preset processing rule to obtain the target text data; and/or determining and screening text data needing to be translated into a target language from the image text recognition result according to a preset processing rule to obtain the target text data.

In some embodiments, the translated target text data and the image data in the commodity picture obtained through the image combination process can be laid in the commodity drawing in a form supporting editing and modification. Therefore, after the merchant obtains the processed commodity picture, the image data in the commodity picture and/or the translated target text data can be edited and modified for the second time according to specific conditions on the basis of the commodity picture, so that the commodity picture is further optimized, and various personalized requirements of the merchant are met.

In some embodiments, the translated target text data and the image data in the commodity picture are respectively arranged in the commodity picture in an independent mode. The merchant can flexibly select one of the translated target text data and the image data in the commodity picture or modify and edit the translated target text data and the image data simultaneously according to specific conditions, so that the merchant can modify and edit the commodity picture more conveniently and flexibly, and personalized requirements of the merchant are further met.

In some embodiments, after obtaining the commodity picture suitable for the target electronic market scene, when the method is implemented, the following may be further included: receiving a modification instruction of a user; and modifying the translated target text data and/or the image data in the commodity picture according to the modification instruction of the user.

In some embodiments, the target picture may specifically also carry address data. E.g., url addresses, etc. The address data can be used for indicating the picture position in the target electronic market scene for putting the commodity picture corresponding to the target picture.

In some embodiments, after obtaining the commodity picture suitable for the target electronic market scene, when the method is implemented, the following may be further included: receiving an issuing instruction of a user; and responding to a user issuing instruction, and replacing the target picture in the target e-commerce scene by using the commodity picture according to the address data.

In this embodiment, in specific implementation, according to address data carried by a target picture, a picture position in a commodity detail page for laying out the commodity picture can be found on a target e-commerce scene, for example, a corresponding shopping website, and then an original target picture at the picture position can be replaced by a processed commodity picture suitable for the target e-commerce scene. Therefore, the operation of the commercial tenant can be simplified, the release of the commodity pictures can be efficiently and conveniently completed, the release efficiency is improved, and the commercial tenant can conveniently and rapidly shop in a target e-commerce scene.

In some embodiments, the target picture may further carry other identification identifiers corresponding to picture positions in the target e-commerce scene for launching the commodity picture corresponding to the target picture, for example, a picture number of the target picture in the target e-commerce scene. Correspondingly, the corresponding picture position can be found in the target e-commerce scene according to the identity of the target picture in response to the issuing instruction of the user, and then the target picture at the picture position is automatically replaced by the commodity picture.

As can be seen from the above, in the processing method of the commodity picture provided in the embodiment of the present specification, the target picture to be translated into the target language is subjected to the preset processing corresponding to the target e-market scene, so as to obtain the target text data matched with the target e-market scene from the target picture; then, the target text data is translated into a target language in a targeted manner to obtain translated target text data; and carrying out image combination processing according to the translated target text data and the target picture to obtain a commodity picture suitable for the target electronic market scene. Therefore, the text data in the target picture can be accurately translated into the target language, and meanwhile, the commodity picture which is suitable for the target e-commerce scene and has a good putting effect is obtained. And before the translation processing, the image text recognition result is correspondingly processed according to the preset processing rule matched with the target E-market scene to obtain target text data which needs to be translated and is suitable for the target E-market scene, so that the subsequent translation processing of text data which does not require translation on the target E-market scene, which wastes processing time and resources, can be avoided, the processing effect is improved, meanwhile, the effect can be relatively better obtained, and the method is more suitable for the translated target text data of the target E-market scene. And after the translated target text data is obtained, the target picture and the translated target text data are subjected to image combination processing according to a preset image combination rule matched with the target E-commerce scene, so that the commodity picture which is suitable for the target E-commerce scene, can better attract consumers of the target E-commerce scene and has a better putting effect can be obtained. And the translated target text data and the image data in the pictures are laid out in a relatively independent form by supporting editing and modification in the commodity pictures obtained through the image combination processing, so that the commodity pictures fed back to the merchants are allowed to be edited and modified according to specific conditions, and the diversified processing requirements of the merchants can be met.

The embodiment of the specification further provides a picture processing method. When the method is implemented, the following contents may be included.

S1: acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to an object and text data related to the object.

S2: and responding to the picture processing request, performing preset processing on the target picture so as to perform preset processing on the target picture from the target picture, and acquiring target text data matched with a target application scene from the target picture.

S3: and translating the target text data into a target language to obtain translated target text data.

S4: and according to the translated target text data and the target picture, carrying out image combination processing to obtain a processed picture suitable for a target scene.

By the method, the target text data needing to be translated in the target picture can be accurately translated into the target language, and the processed picture suitable for the target application scene can be obtained.

The embodiment of the specification also provides another commodity picture processing method. The method can be applied to the terminal equipment side. In particular implementations, the method may include the following.

S1: displaying a picture processing request interface, and receiving a target picture and an indication parameter of a target language through the picture processing request interface, wherein the target picture comprises image data related to a commodity and text data related to the commodity.

S2: generating and sending a picture processing request to a server according to the target picture and the indication parameters of the target language, wherein the server is used for responding to the picture processing request to perform preset processing on the target picture so as to acquire target text data matched with a target electronic scene using the target language from the target picture; translating the target text data into a target language to obtain translated target text data; and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

S3: and receiving and displaying the commodity picture.

In some embodiments, after receiving and displaying the commodity picture, when the method is implemented, the following may be further included: displaying a picture modification request interface, and receiving a modification instruction aiming at translated target text data and/or image data in the commodity picture through the picture modification request interface; and modifying the translated target text data and/or the image data in the commodity picture according to the modification instruction.

By the mode, the merchant can translate the target picture into the target language efficiently and conveniently, and obtain the commodity picture which is suitable for the target electronic market scene and has a better putting effect. Furthermore, the commodity picture can be modified and edited based on the commodity picture according to own ideas, so that the commodity picture is further optimized in a targeted manner, and the diversified processing requirements of merchants can be better met.

In some embodiments, the picture modification request interface may specifically include: the image preview area, the image layer area, the preview area, the edit area and other operation areas. Of course, the above listed operation regions are only a schematic illustration. In specific implementation, according to a specific application scenario, other types of operation areas may be further set in the picture modification request interface, so as to facilitate the operation of the merchant. The present specification is not limited to these.

The embodiment of the present specification further provides another method for processing a commodity picture, and the specific implementation may include the following contents.

S1: and acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to the commodity and text data related to the commodity.

S2: and obtaining a commodity picture suitable for the target electronic market scene according to the target picture and the indication parameters of the target language, wherein the commodity picture suitable for the target electronic market scene comprises the steps of translating target text data matched with the target electronic market scene in the target picture into the target language, and processing the obtained commodity picture through image combination.

S3: and receiving and replacing the target picture in the target E-commerce scene by using the commodity picture suitable for the target E-commerce scene according to the issuing instruction fed back by the user aiming at the commodity picture suitable for the target E-commerce scene.

Therefore, the operation of the commercial tenant can be simplified, the target picture required to be processed by the commercial tenant can be efficiently processed into the commodity picture suitable for the target E-commerce scene, the commercial tenant instruction can be responded in time, the originally used target picture in the target E-commerce scene is replaced by the processed commodity picture, and the release of the commodity picture is rapidly completed.

The embodiment of the present specification further provides another picture processing method, which may include the following contents in specific implementation.

S1: and receiving and displaying the target picture sent by the second user to the first user.

S2: and acquiring an indication parameter of a target language of the first user for the target picture.

S3: generating and sending a picture processing request aiming at the target picture to a server according to the indication parameter of the target language, wherein the server is used for responding to the picture processing request and carrying out preset processing on the target picture so as to obtain target text data from the target picture; translating the target text data into a target language to obtain translated target text data; and carrying out image combination processing according to the translated target text data and the target image to obtain a processed image suitable for the first user.

S4: and acquiring and displaying the processed picture to the first user.

In this embodiment, the method may be specifically applied to a terminal device used by a first user.

In this embodiment, when a first user uses a terminal device to interact with a second user, a target picture based on other languages sent by the second user to the first user is often received.

For example, when a first user chats with a second user located abroad by using a chat APP on a terminal device, the second user may send a picture (as a target picture) containing foreign language, which is taken abroad by the second user, to the first user. The photo may include text data based on a foreign language. For example, a photograph of a cosmetic product including a foreign language label taken by the second user in a foreign supermarket, or the like.

In this embodiment, the terminal device receives the target picture and displays the target picture to the first user. If the first user wants to translate the text data in the target picture into the target language suitable for the first user, the first user can instruct the terminal device to process the target picture by performing corresponding operation on the terminal device, and set an indicating parameter suitable for the target language.

The terminal device can obtain an indication parameter of a target language of the first user for the target picture. And generating and sending a picture processing request aiming at the target picture to a server according to the indication parameter of the target language. The picture processing request may carry a target picture and an indication parameter of a target language set by the first user.

The server receives and responds to the picture processing request, and performs preset processing on the target picture to acquire target text data from the target picture; translating the target text data into a target language to obtain translated target text data; and carrying out image combination processing according to the translated target text data and the target image to obtain a processed image suitable for the first user, and feeding back the processed image suitable for the first user to the terminal equipment.

And the terminal equipment receives and displays the processed picture suitable for the first user to the first user. Thereby a relatively better usage experience for the first user may be achieved.

In some embodiments, the target picture may specifically include at least one of: pictures containing recommended goods, pictures containing promotional information, pictures containing notification content, and the like. Of course, the above listed target pictures are only illustrative. In a specific implementation, the target picture may further include other types or contents of pictures according to specific situations. The present specification is not limited to these.

S1: and acquiring a target picture, wherein the target picture is a picture which is indicated by a second user and sent to a first user.

S2: determining whether the language in the target picture matches a first user.

S3: and under the condition that the language in the target picture is determined to be not matched with the first user, determining the language used by the first user as the target language, and processing the target picture based on the target language to obtain a processed picture suitable for the first user.

S4: and sending the processed picture to the first user.

In some embodiments, the above method may be particularly applicable to a server of a data processing system. The data processing system may be a background data processing system of a chat APP, a mailbox, or other websites or applications supporting picture data transmission.

In some embodiments, when the second user sends the target picture to the first user through the chat APP, the mailbox, or another website or application program supporting picture data transmission, the server may first obtain the target picture that the second user indicates to send to the first user, and detect and determine whether the language in the target picture matches with the first user.

For example, the region where the first user is located may be obtained and determined according to the geographical location of the first user, and the language used by the region where the first user is located is used as the target language matched with the first user. Meanwhile, whether the language used by the text data contained in the target picture is the same as the target language is detected. And if the target picture is the same as the first user, determining that the language in the target picture is matched with the first user. And if not, determining that the language in the target picture does not match the first user.

Of course, the above listed manner of determining whether the language in the target picture matches the first user is only an illustrative example. In specific implementation, according to specific situations, other suitable manners may also be adopted to determine whether the language in the target picture matches the first user. The present specification is not limited to these.

When it is determined that the language in the target picture is not matched with the first user, the language used by the first user may be determined as the target language, and the target picture is processed based on the target language to obtain a processed picture suitable for the first user.

And the subsequent server can directly send the processed picture suitable for the first user to the first user, or send the processed picture suitable for the first user to the first user together with the original target picture.

Therefore, after the first user receives the picture, the text content contained in the target picture sent by the second user can be conveniently and directly read and understood, and the use experience of the first user is further improved.

In some embodiments, the processing of the target picture based on the target language may include the following steps: presetting the target picture to acquire target text data from the target picture; translating the target text data into a target language to obtain translated target text data; and carrying out image combination processing according to the translated target text data and the target image to obtain a processed image suitable for the first user.

Embodiments of the present specification further provide a server, including a processor and a memory for storing processor-executable instructions, where the processor, when implemented, may perform the following steps according to the instructions: acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to commodities and text data related to the commodities; responding to the picture processing request, and performing preset processing on the target picture to acquire target text data matched with a target electronic market using a target language from the target picture; translating the target text data into a target language to obtain translated target text data; and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

In order to more accurately complete the above instructions, referring to fig. 11, another specific server is provided in the embodiments of the present specification, wherein the server includes a network communication port 1101, a processor 1102 and a memory 1103, and the above structures are connected by an internal cable, so that the structures can perform specific data interaction.

The network communication port 1101 may be specifically configured to acquire a target picture and an indication parameter of a target language, where the target picture includes image data related to a product and text data related to the product.

The processor 1102 may be specifically configured to perform preset processing on the target picture, so as to obtain target text data matched with a target electronic market using a target language from the target picture; translating the target text data into a target language to obtain translated target text data; and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

The memory 1103 may be specifically configured to store a corresponding instruction program.

In this embodiment, the network communication port 1101 may be a virtual port bound to different communication protocols so as to transmit or receive different data. For example, the network communication port may be port No. 80 responsible for web data communication, port No. 21 responsible for FTP data communication, or port No. 25 responsible for mail data communication. In addition, the network communication port can also be a communication interface or a communication chip of an entity. For example, it may be a wireless mobile network communication chip, such as GSM, CDMA, etc.; it can also be a Wifi chip; it may also be a bluetooth chip.

In this embodiment, the processor 1102 may be implemented in any suitable manner. For example, a processor may take the form of, for example, a microprocessor or processor and a computer-readable medium that stores computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, an embedded microcontroller, and so forth. The description is not intended to be limiting.

In this embodiment, the memory 1103 may include multiple layers, and in a digital system, the memory may be any memory as long as it can store binary data; in an integrated circuit, a circuit without a physical form and with a storage function is also called a memory, such as a RAM, a FIFO and the like; in the system, the storage device in physical form is also called a memory, such as a memory bank, a TF card and the like.

An embodiment of the present specification further provides a terminal device, including a processor and a memory for storing processor-executable instructions, where the processor, when implemented specifically, may perform the following steps according to the instructions: displaying a picture processing request interface, and receiving a target picture and an indication parameter of a target language through the picture processing request interface, wherein the target picture comprises image data related to a commodity and text data related to the commodity; generating and sending a picture processing request to a server according to the target picture and the indication parameters of the target language, wherein the server is used for responding to the picture processing request to perform preset processing on the target picture so as to acquire target text data matched with a target electronic scene using the target language from the target picture; translating the target text data into a target language to obtain translated target text data; according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene; and receiving and displaying the commodity picture.

The embodiment of the present specification further provides a computer storage medium based on the processing method of the commodity picture, where the computer storage medium stores computer program instructions, and when the computer program instructions are executed, the computer storage medium implements: acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to commodities and text data related to the commodities; responding to the picture processing request, and performing preset processing on the target picture to acquire target text data matched with a target electronic market using a target language from the target picture; translating the target text data into a target language to obtain translated target text data; and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

In this embodiment, the storage medium includes, but is not limited to, a Random Access Memory (RAM), a Read-only memory (ROM), a Cache (Cache), a hard disk (hard disk drive, HDD), or a memory card (memory card). The memory may be used to store computer program instructions. The network communication unit may be an interface for performing network connection communication, which is set in accordance with a standard prescribed by a communication protocol.

In this embodiment, the functions and effects specifically realized by the program instructions stored in the computer storage medium can be explained by comparing with other embodiments, and are not described herein again.

Referring to fig. 12, in a software level, an embodiment of the present specification further provides a device for processing a commodity picture, where the device may specifically include the following structural modules.

The receiving module 1201 may be specifically configured to acquire a target picture and an indication parameter of a target language, where the target picture includes image data related to a product and text data related to the product.

The preset processing module 1202 may be specifically configured to perform preset processing on the target picture, so as to obtain target text data matched with a target electronic market using a target language from the target picture.

The translation processing module 1203 may be specifically configured to translate the target text data into a target language, so as to obtain translated target text data.

The image combination processing module 1204 may be specifically configured to perform image combination processing according to the translated target text data and the target picture to obtain a commodity picture suitable for the target e-market scene.

In some embodiments, the default processing module 1202 may specifically include the following structural units.

The identification unit may be specifically configured to perform image text identification on the target picture to obtain an image text identification result.

The processing unit may be specifically configured to perform corresponding processing on the image text recognition result according to a preset processing rule to obtain the target text data, where the preset processing rule includes a processing rule matched with a target electronic market.

In some embodiments, the image text recognition result may specifically include one or more of the following data: text data in the target picture, position coordinates of the text data in the target picture, background color of the text data in the target picture, character features of the text data, and the like.

In some embodiments, when the processing unit is implemented, the processing unit may determine and remove text data that does not need to be translated into a target language from the image text recognition result according to a preset processing rule to obtain the target text data; and/or determining and screening text data needing to be translated into a target language from the image text recognition result according to a preset processing rule to obtain the target text data.

In some embodiments, when the translation processing module 1203 is implemented specifically, a preset translation model may be called, and the target text data is translated into a target language to obtain the translated target text data, where the preset translation model includes a translation model obtained by learning text data in a target e-commerce scene in advance.

In some embodiments, when the matching processing module 1204 is specifically implemented, the target picture and the translated target text data may be subjected to matching processing according to a preset matching rule, so as to obtain the commodity picture suitable for the target e-commerce scene, where the preset matching rule includes a typesetting rule matched with the target e-commerce scene.

It should be noted that, the units, devices, modules, etc. illustrated in the above embodiments may be implemented by a computer chip or an entity, or implemented by a product with certain functions. For convenience of description, the above devices are described as being divided into various modules by functions, and are described separately. It is to be understood that, in implementing the present specification, functions of each module may be implemented in one or more pieces of software and/or hardware, or a module that implements the same function may be implemented by a combination of a plurality of sub-modules or sub-units, or the like. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

As can be seen from the above, in the processing apparatus for commodity pictures provided in the embodiments of the present specification, a preset processing module is used to perform preset processing corresponding to a target e-market scene on a target picture to be translated into a target language, so as to obtain target text data matched with the target e-market scene from the target picture; the target text data is translated into a target language through a translation processing module in a targeted manner, and translated target text data is obtained; and carrying out image combination processing according to the translated target text data and the target picture through an image combination processing module to obtain a commodity picture suitable for the target electronic market scene. Therefore, the text data in the target picture can be accurately translated into the target language, and the commodity picture which is suitable for the target e-commerce scene and has a good putting effect can be obtained.

Although the present specification provides method steps as described in the examples or flowcharts, additional or fewer steps may be included based on conventional or non-inventive means. The order of steps recited in the embodiments is merely one manner of performing the steps in a multitude of orders and does not represent the only order of execution. When an apparatus or client product in practice executes, it may execute sequentially or in parallel (e.g., in a parallel processor or multithreaded processing environment, or even in a distributed data processing environment) according to the embodiments or methods shown in the figures. The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, the presence of additional identical or equivalent elements in a process, method, article, or apparatus that comprises the recited elements is not excluded. The terms first, second, etc. are used to denote names, but not any particular order.

Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may therefore be considered as a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

This description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, classes, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

From the above description of the embodiments, it is clear to those skilled in the art that the present specification can be implemented by software plus necessary general hardware platform. With this understanding, the technical solutions in the present specification may be essentially embodied in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which may be a personal computer, a mobile terminal, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments in the present specification.

The embodiments in the present specification are described in a progressive manner, and the same or similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. The description is operational with numerous general purpose or special purpose computing system environments or configurations. For example: personal computers, server computers, hand-held or portable devices, tablet-type devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.

While the specification has been described with examples, those skilled in the art will appreciate that there are numerous variations and permutations of the specification that do not depart from the spirit of the specification, and it is intended that the appended claims include such variations and modifications that do not depart from the spirit of the specification.

Claims

1. A commodity picture processing method comprises the following steps:

acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to commodities and text data related to the commodities;

presetting the target picture to acquire target text data matched with a target electronic market using a target language from the target picture;

translating the target text data into a target language to obtain translated target text data;

and according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene.

2. The method of claim 1, wherein the pre-processing the target picture to obtain target text data from the target picture that matches a target e-market scene in a target language comprises:

performing image text recognition on the target picture to obtain an image text recognition result;

and correspondingly processing the image text recognition result according to preset processing rules to obtain the target text data, wherein the preset processing rules comprise processing rules matched with the target E-market scene.

3. The method of claim 2, the image text recognition results comprising one or more of the following: text data in the target picture, position coordinates of the text data in the target picture, background color of the text data in the target picture, and character features of the text data.

4. The method according to claim 3, wherein the correspondingly processing the image text recognition result according to a preset processing rule to obtain the target text data comprises:

determining and removing text data which does not need to be translated into a target language from the image text recognition result according to a preset processing rule to obtain the target text data;

and/or the presence of a gas in the gas,

and determining and screening text data needing to be translated into a target language from the image text recognition result according to a preset processing rule to obtain the target text data.

5. The method according to claim 4, wherein the processing the image text recognition result according to a preset processing rule includes:

determining text characteristics of text data in an image text recognition result;

screening out text data with text characteristics conforming to first-class text characteristics from the image text recognition result as first-class text data; wherein the first type of text data comprises: the text data is arranged on the commodity body, and/or the text data is arranged on the trademark of the commodity;

and removing the first type of text data from the image text recognition result.

6. The method according to claim 3, wherein the processing the image text recognition result according to a preset processing rule includes:

determining multi-line text data belonging to the same text region according to the position coordinates of the text data in the image text recognition result;

plural lines of text data belonging to the same text area are merged into one text data.

7. The method of claim 4, wherein the character characteristics of the text data include an arrangement direction of characters in the text data;

correspondingly, the correspondingly processing the image text recognition result according to the preset processing rule includes:

screening out text data with vertically arranged characters from the image text recognition result as second type text data according to character features of the text data;

and converting the second type text data into text data arranged along a preset character arrangement direction according to the character arrangement direction in the second type text data, wherein the text data is used as the processed second type text data, and vertical arrangement labels are arranged in the processed second type text data.

8. The method according to claim 4, wherein the processing the image text recognition result according to a preset processing rule includes:

screening out text data containing unit words from the image text recognition result through unit word retrieval to serve as third type text data;

determining whether the unit words in the third type of text data are matched with a target E-market scene;

under the condition that the unit words in the third type of text data are determined to be not matched with the target E-market scenery, determining matched unit words corresponding to the unit words;

and modifying the unit words in the third type text data and the numerical characters adjacent to the unit words according to the conversion rule of the matched unit words to obtain the processed third type text data.

9. The method of claim 8, the unit word comprising one or more of: length unit words, weight unit words, size units, currency unit words.

10. The method according to claim 3, wherein the processing the image text recognition result according to a preset processing rule includes:

determining a target culture sensitive rule matched with a target E-market scene;

and correspondingly modifying the data in the image text recognition result according to the target culture sensitivity rule.

11. The method according to claim 4, wherein the processing the image text recognition result according to a preset processing rule includes:

according to a preset filtering rule, fourth type text data are screened out from the image text recognition result, wherein the fourth type text data comprise at least one of the following text data: the method comprises the following steps of (1) based on favorable data of an initial e-commerce scene, service description aiming at the initial e-commerce scene, and a promotion scheme aiming at the initial e-commerce scene;

and removing the fourth type of text data from the image text recognition result.

12. The method of claim 1, said translating said target text data into a target language resulting in translated target text data, comprising:

and calling a preset translation model, translating the target text data into a target language, and obtaining the translated target text data, wherein the preset translation model comprises a translation model obtained by learning the text data in the target e-commerce scene in advance.

13. The method of claim 3, wherein the combining process is performed according to the translated target text data and the target picture to obtain a commodity picture suitable for the target electronic market scene, and the method comprises:

and carrying out image combination processing on the target picture and the translated target text data according to preset image combination rules to obtain the commodity picture suitable for the target E-commerce scene, wherein the preset image combination rules comprise typesetting rules matched with the target E-commerce scene.

14. The method according to claim 13, wherein the matching the target picture and the translated target text data according to a preset matching rule includes:

determining a picture area where the target text data in the target picture is located according to the position coordinates of the target text data;

covering a picture area where the target text data is located by using a background color block with the same color as the background color of the target text data;

and laying the translated target text data above the background color blocks.

15. The method of claim 14, said laying out translated target text data, comprising:

determining character characteristics of the translated target text data according to the character characteristics of the target text data;

and laying the translated target text data according to the character characteristics of the translated target text data.

16. The method of claim 15, further comprising:

determining the area size of an image area where target text data in a target image are located and the number of characters contained in the translated target text data;

and adjusting character characteristics of the translated target text data according to the area size of the picture area where the target text data is located and the number of characters contained in the translated target text data.

17. The method of claim 14, said laying out translated target text data, further comprising:

detecting whether the translated target text data is provided with a vertical arrangement label;

under the condition that the translated target text data is determined to be provided with the vertically arranged tags, rotating the translated target text data by 90 degrees in the clockwise direction to obtain the rotated target text data;

and laying the rotated target text data.

18. The method of claim 1, after obtaining the picture of the item suitable for the target electronic marketplace, the method further comprising:

receiving a modification instruction of a user;

and modifying the translated target text data and/or the image data in the commodity picture according to the modification instruction of the user.

19. The method of claim 1, wherein the target picture further carries address data.

20. The method of claim 19, after obtaining the picture of the item suitable for the target electronic marketplace, further comprising:

receiving an issuing instruction of a user;

and responding to a user issuing instruction, and replacing the target picture in the target e-commerce scene by using the commodity picture according to the address data.

21. A picture processing method comprises the following steps:

acquiring a target picture and an indication parameter of a target language, wherein the target picture comprises image data related to an object and text data related to the object;

presetting the target picture to acquire target text data matched with a target application scene from the target picture;

and according to the translated target text data and the target picture, carrying out image combination processing to obtain a processed picture suitable for a target scene.

22. A commodity picture processing method comprises the following steps:

displaying a picture processing request interface, and receiving a target picture and an indication parameter of a target language through the picture processing request interface, wherein the target picture comprises image data related to a commodity and text data related to the commodity;

generating and sending a picture processing request to a server according to the target picture and the indication parameters of the target language, wherein the server is used for responding to the picture processing request to perform preset processing on the target picture so as to acquire target text data matched with a target electronic scene using the target language from the target picture; translating the target text data into a target language to obtain translated target text data; according to the translated target text data and the target picture, carrying out image combination processing to obtain a commodity picture suitable for the target electronic market scene;

and receiving and displaying the commodity picture.

23. The method of claim 22, after receiving and displaying the picture of the merchandise, the method further comprising:

displaying a picture modification request interface, and receiving a modification instruction aiming at translated target text data and/or image data in the commodity picture through the picture modification request interface;

and modifying the translated target text data and/or the image data in the commodity picture according to the modification instruction.

24. A commodity picture processing method comprises the following steps:

obtaining a commodity picture suitable for a target electronic market scene according to the target picture and the indication parameters of the target language, wherein the commodity picture suitable for the target electronic market scene comprises the steps of translating target text data matched with the target electronic market scene in the target picture into the target language, and processing the obtained commodity picture through image combination;

and receiving and replacing the target picture in the target E-commerce scene by using the commodity picture suitable for the target E-commerce scene according to the issuing instruction fed back by the user aiming at the commodity picture suitable for the target E-commerce scene.

25. A picture processing method comprises the following steps:

receiving and displaying a target picture sent by a second user to a first user;

acquiring an indication parameter of a target language of a first user for the target picture;

generating and sending a picture processing request aiming at the target picture to a server according to the indication parameter of the target language, wherein the server is used for responding to the picture processing request and carrying out preset processing on the target picture so as to obtain target text data from the target picture; translating the target text data into a target language to obtain translated target text data; carrying out image combination processing according to the translated target text data and the target image to obtain a processed image suitable for a first user;

and acquiring and displaying the processed picture to the first user.

26. The method of claim 25, wherein the target picture comprises at least one of: the picture containing the recommended commodity, the picture containing the promotion information and the picture containing the notification content.

27. A picture processing method comprises the following steps:

acquiring a target picture, wherein the target picture is a picture which is indicated by a second user and sent to a first user;

determining whether the language in the target picture matches a first user;

under the condition that the language in the target picture is determined to be not matched with the first user, determining the language used by the first user as the target language, and processing the target picture based on the target language to obtain a processed picture suitable for the first user;

and sending the processed picture to the first user.

28. The method of claim 27, processing the target picture based on the target language, comprising:

presetting the target picture to acquire target text data from the target picture;

and carrying out image combination processing according to the translated target text data and the target image to obtain a processed image suitable for the first user.

29. A device for processing a picture of an article, comprising:

the system comprises a receiving module, a processing module and a display module, wherein the receiving module is used for acquiring a target picture and an indication parameter of a target language, and the target picture comprises image data related to commodities and text data related to the commodities;

the preset processing module is used for carrying out preset processing on the target picture so as to acquire target text data matched with a target electronic market using a target language from the target picture;

the translation processing module is used for translating the target text data into a target language to obtain translated target text data;

and the image combination processing module is used for carrying out image combination processing according to the translated target text data and the target image to obtain a commodity image suitable for the target electronic market scene.

30. A server comprising a processor and a memory for storing processor-executable instructions which, when executed by the processor, implement the steps of the method of any one of claims 1 to 20.

31. A terminal device comprising a processor and a memory for storing processor-executable instructions which, when executed by the processor, implement the steps of the method of any of claims 22 to 23.

32. A computer readable storage medium having stored thereon computer instructions which, when executed, implement the steps of the method of any one of claims 1 to 20.