WO2021169384A1 - 信息识别方法及装置、系统、电子设备、存储介质和计算机程序 - Google Patents

信息识别方法及装置、系统、电子设备、存储介质和计算机程序 Download PDF

Info

Publication number
WO2021169384A1
WO2021169384A1 PCT/CN2020/126167 CN2020126167W WO2021169384A1 WO 2021169384 A1 WO2021169384 A1 WO 2021169384A1 CN 2020126167 W CN2020126167 W CN 2020126167W WO 2021169384 A1 WO2021169384 A1 WO 2021169384A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
information
recognized
identified
recognition
Prior art date
Application number
PCT/CN2020/126167
Other languages
English (en)
French (fr)
Inventor
韩旭
Original Assignee
北京市商汤科技开发有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京市商汤科技开发有限公司 filed Critical 北京市商汤科技开发有限公司
Priority to KR1020217011452A priority Critical patent/KR20210110562A/ko
Priority to JP2021520536A priority patent/JP2022524672A/ja
Publication of WO2021169384A1 publication Critical patent/WO2021169384A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/49—Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects

Definitions

  • the present disclosure relates to the field of computer vision, and in particular to an information recognition method and device, system, electronic equipment, storage medium, and computer program.
  • the present disclosure proposes an information identification scheme.
  • an information recognition method including: a first device obtains an image sequence of an object to be recognized, the image sequence includes at least two frames of images; Determine at least one frame of image as the image to be identified; determine the information to be identified according to the image to be identified; send the information to be identified to the second device, so that the second device obtains the first image based on the information to be identified Recognition results.
  • the first device determines at least one frame of image from the image sequence as the image to be recognized, including: the first device according to the recognition state and/or image of the image sequence Quality, at least one frame of image is determined from the image sequence as the image to be recognized.
  • the first device selects at least one frame of image from the image sequence as the image to be recognized according to the recognition state of the image sequence, including: At least part of the images in the sequence of images are respectively identified by second information to obtain one or more second recognition results; the first device obtains a reference recognition result according to the one or more second recognition results; each of the The second recognition result is compared with the reference recognition result to obtain the confidence level of each second recognition result; the first device determines each image corresponding to at least a part of all the second recognition results according to the confidence level Is the image to be recognized, wherein the determined confidence of the image to be recognized is higher than the confidence of the image not to be recognized in the image sequence.
  • the first device selects at least one frame of image from the image sequence as the image to be recognized according to the image quality of the image sequence, including: the first device separately obtains the Image quality of at least part of the images in the image sequence in at least one measurement dimension; the first device determines at least one frame of images in the image sequence whose image quality is greater than a threshold value in the corresponding measurement dimension as the image to be recognized.
  • the measurement dimension includes at least one of sharpness, integrity, strong light conditions, dark light conditions, and occlusion conditions.
  • the method further includes: saving at least one frame in the image to be recognized.
  • the first device to determine the information to be identified according to the image to be identified includes: the first device performs first encryption processing and/or signature processing on the image to be identified, Get the information to be identified.
  • the first encryption processing includes: encoding and encrypting the image to be recognized to obtain first encrypted information;
  • the signature processing includes: adding signature information of the first device to The image to be recognized.
  • the object to be recognized includes a card object and/or a form object; the first recognition result includes at least one of a text, a logo, and a picture recorded in the object to be recognized.
  • an information identification method including: a second device receives information to be identified of an object to be identified; the second device performs first information identification on the information to be identified to obtain the first information Identification result; the second device sends identification information to the first device according to the first identification result.
  • the second device performs first information recognition on the to-be-identified information to obtain the first recognition result, including: the second device obtains the to-be-recognized image included in the to-be-identified information Perform anti-counterfeiting detection on the image to be identified to obtain a detection result; in the case where the detection result is passed, the second device performs first information identification on the image to be identified to obtain a first identification result.
  • the acquiring, by the second device, the image to be identified included in the information to be identified includes: acquiring, by the second device, the signature information included in the information to be identified; When the information matches the signature information of the first device, the second device decrypts the first encrypted information included in the information to be identified to obtain the image to be identified.
  • the second device performs anti-counterfeiting detection on the image to be recognized to obtain the detection result, including: the second device classifies the image to be recognized to obtain the image to be recognized In the case where the classification result indicates that the image to be identified is an image obtained by shooting the object to be identified, the second device records the detection result as passed; and/or In a case where the classification result indicates that the image to be identified is an image obtained by photographing a copy or a remake of the object to be identified, the second device records the detection result as a failure.
  • the second device performs first information recognition on the image to be recognized to obtain the first recognition result, including: the second device uses an optical character recognition OCR model to recognize the to-be-recognized image.
  • the recognition image is subjected to OCR recognition, and the first recognition result is obtained.
  • the second device performs first information recognition on the to-be-identified information to obtain the first recognition result
  • the method further includes: the second device judging whether the first recognition result is consistent with the predetermined Set the rule to match, and get the judgment result.
  • the second device sending identification information to the first device according to the first identification result includes: the second device performs a second encryption process on the first identification result, Or perform the second encryption processing on the first identification result and the judgment result to obtain second encrypted information; the second device uses the second encrypted information as the identification information and sends it to the The first device sends.
  • the method further includes: the second device storing at least one frame in the image to be recognized.
  • an information recognition device including: an image sequence acquisition module for acquiring an image sequence of an object to be recognized, the image sequence including at least two frames of images; At least one frame of image is determined in the image sequence as the image to be identified; the information to be identified generating module is used to determine the information to be identified based on the image to be identified; the information to be identified sending module is used to send the image to the second device Information to be identified, so that the second device obtains the first identification result according to the information to be identified.
  • an information identification device including: a receiving module for receiving information to be identified of an object to be identified; and an identification module for performing first information identification on the information to be identified to obtain The first identification result; an identification information sending module, configured to send identification information to the first device according to the first identification result.
  • an electronic device including: a processor; a non-transitory storage medium for storing instructions executable by the processor; wherein the processor is configured to: execute the above-mentioned first aspect The method of information identification.
  • an electronic device including: a processor; a non-transitory storage medium for storing instructions executable by the processor; wherein the processor is configured to execute the above-mentioned second aspect The method of information identification.
  • a computer-readable storage medium having computer program instructions stored thereon, and when the computer program instructions are executed by a processor, the above-mentioned information identification method of the first aspect is realized.
  • a computer-readable storage medium having computer program instructions stored thereon, and the computer program instructions, when executed by a processor, implement the information recognition method of the second aspect described above.
  • a ninth aspect of the present disclosure there is provided a computer program that, when executed by a processor, implements the information identification method of any one of the above aspects.
  • the information to be identified is sent to the second device according to the image to be identified, so that the second device
  • the first recognition result is obtained according to the information to be recognized.
  • Fig. 1 shows a flowchart of an information recognition method according to an embodiment of the present disclosure.
  • Fig. 2 shows a flowchart of an information recognition method according to an embodiment of the present disclosure.
  • Fig. 3 shows a block diagram of an information recognition device according to an embodiment of the present disclosure.
  • Fig. 4 shows a block diagram of an information recognition device according to an embodiment of the present disclosure.
  • Fig. 5 shows a schematic diagram of an application example according to the present disclosure.
  • Fig. 6 shows a block diagram of an electronic device according to an embodiment of the present disclosure.
  • Fig. 7 shows a block diagram of an electronic device according to an embodiment of the present disclosure.
  • Fig. 1 shows a flow chart of an information recognition method according to an embodiment of the present disclosure.
  • the method can be applied to a first device.
  • the first device can be a device that can collect an image sequence of an object to be identified, that is, a device with an information collection function.
  • the first device may be a hardware device such as a terminal device with a shooting function or other processing devices.
  • terminal devices can be User Equipment (UE), mobile devices, user terminals, terminals, cellular phones, cordless phones, personal digital assistants (PDAs), handheld devices, computing devices, vehicle-mounted devices, and portable devices. Wearable equipment, etc.
  • UE User Equipment
  • PDAs personal digital assistants
  • the operation of collecting the image sequence of the object to be identified can also be performed by a separate front-end device such as a camera, and then the front-end device sends the collected image sequence of the object to be identified to the first device. Then, the information identification method is implemented by the processor of the first device invoking the computer-readable instructions stored in the memory.
  • the information identification method can be applied to the first device, and includes the following steps.
  • Step S11 Obtain an image sequence of the object to be identified, where the image sequence includes at least two frames of images.
  • Step S12 Determine at least one frame of image from the image sequence as the image to be recognized.
  • Step S13 Determine the information to be identified according to the image to be identified.
  • Step S14 Send the information to be identified to the second device, so that the second device obtains the first identification result according to the information to be identified.
  • the object to be identified is an object to be identified.
  • the object to be identified can have multiple manifestations.
  • the object to be identified may be an object with uniform specifications such as a card object or a form object.
  • the card object can be an ID card, a pass, a bank card, etc.
  • the form object can be a form of a prescribed standard.
  • the image sequence of the object to be identified may be an image sequence or a group of images obtained by the first device through image acquisition of the object to be identified.
  • the first device may obtain the image sequence of the object to be identified by continuously scanning or video recording the object to be identified.
  • the first device may also acquire multiple frames of images of the object to be identified by taking pictures of the object to be identified at a certain frequency. The acquired multiple frames of images can be formed into an image sequence in the order of acquisition time.
  • the multiple frames of images collected for the object to be identified may not have time correlation with each other. In this case, a group of images may be formed instead of an image sequence.
  • the image to be recognized may be one or more frames selected by the first device from the acquired image sequence.
  • only one frame of higher quality image can be selected from the image sequence as the image to be recognized.
  • multiple frames of images with higher quality can be selected from the image sequence as the image to be recognized.
  • the number of selected images to be processed can be flexibly determined according to the actual situation.
  • a case where one frame of image is selected as the image to be recognized will be described as an example.
  • the implementation principle is similar to the case of selecting one frame of image, and no detailed description will be given.
  • the case of selecting multiple frames of images as the image to be recognized is also included in the scope of the disclosure.
  • the information to be recognized may be determined based on the image to be recognized.
  • the to-be-identified information may be information that contains content that needs to be identified in the to-be-identified object.
  • the to-be-identified information may directly be the selected image to-be-identified.
  • the information to be identified may also be information obtained after certain processing is performed on the image to be identified. For the determination method of the information to be identified, refer to the subsequent disclosed embodiments.
  • the information to be identified may be sent to the second device, so that the second device obtains the first identification result according to the information to be identified.
  • the second device may be a device with an information recognition function.
  • the second device can be implemented by hardware or software.
  • the second device may be a hardware device such as a terminal device, a server, or other processing devices.
  • the terminal device reference may be made to the above disclosed embodiments, which will not be repeated here.
  • the second device when the second device is a server, the second device may be a cloud server or a local server.
  • the functions of the second device may be implemented by a processor such as a CPU reading a computer program in a storage medium.
  • the first recognition result may be the result obtained by the second device according to the information to be recognized, and the specific type and content of the first recognition result may be different according to the expression mode of the object to be recognized, which is not limited in the embodiment of the present disclosure.
  • the object to be recognized may include a card object and/or a form object
  • the first recognition result may include at least one of text, characters, icons, logos, and pictures recorded in the object to be recognized . Since the to-be-identified object, such as the card object or the form object, has a certain format specification, the first identification can be completed based on the standardized content.
  • card objects can include but are not limited to ID cards, bank cards, and passes; form objects can include, but are not limited to, insurance policies and invoices.
  • the first recognition result obtained can often be based on actual needs and the corresponding standardized content of the object to be recognized, and there is an adjustment of the category.
  • the first identification result may include one or more of the name, date of birth, place of residence, ID number, portrait, etc. recorded on the front of the ID card.
  • the first recognition result may also be a further recognition result obtained based on the foregoing content, for example, by further recognizing the recognized ID card number to determine the birthplace of the person to whom the ID card belongs and other data.
  • the first recognition result may be the content recorded in the object to be recognized, such as recorded text, pictures, or an identifier used to indicate the category of the object to be recognized.
  • the first recognition result may be the text contained in the card object, such as the address on the ID card, the bank card number on the bank card, and so on. Subsequent disclosed embodiments are explained by taking the first recognition result as text as an example, and the first recognition result is a picture or a logo or other types of situations, which can be flexibly extended according to the subsequent disclosed embodiments, and no detailed description will be provided.
  • the specific implementation manner can refer to the subsequent disclosed embodiments.
  • the information to be recognized is determined according to the selected image to be recognized and sent to the second device The information to be identified, so that the second device obtains the first identification result according to the information to be identified.
  • the object to be identified is first automatically selected, and then the second device is used for identification. Because the automatic frame selection can select the image with the best image quality, the image used for recognition has higher image accuracy and recognition effect than the directly acquired image, so the final first recognition result can be more accurate, and at the same time It can also reduce the failure rate of the information recognition process, thereby improving the user experience in the information recognition process.
  • step S12 may include:
  • At least one frame of image is selected as the image to be recognized.
  • the recognition state may indicate the state of the image content in the recognized image sequence, for example, the degree of confidence of the recognized image content.
  • the better the recognition status of the image content the more comprehensive and accurate the recognizable content contained in the image. That is, the process of determining the image to be recognized according to the recognition state can be understood as determining the image to be recognized according to the pros and cons of the recognition state between different frame images in the image sequence.
  • the first recognition result obtained based on the image to be recognized can be more comprehensive and have higher accuracy.
  • the image quality can be the quality of each frame of the image sequence, and the image quality can be judged by a preset quality index.
  • the quality index may include one or more of sharpness, completeness, strong light conditions, dark light conditions, and occlusion conditions, which are not limited herein.
  • the higher the image quality the easier it is to recognize the frame of the image, and the recognition result is more accurate. That is, the process of determining the image to be recognized according to the image quality can be understood as determining the image to be recognized according to the level of image quality between different frame images of the image sequence.
  • the first recognition result obtained based on the image to be recognized can also be more accurate and has a better recognition effect.
  • the specific implementation method for determining the image to be recognized according to the recognition state or the image quality can also be flexibly selected according to the actual situation, and you can refer to the subsequent disclosed embodiments.
  • the determination of the image to be recognized from the image sequence may include two methods: determination based on the recognition state of the image sequence and determination based on the image quality of the image sequence.
  • it may be determined only according to the recognition state or the image quality, that is, the first device automatically determines the image to be recognized in one of these methods.
  • it can also be determined based on the recognition state and the image quality at the same time, that is, the first device can determine the image to be recognized by combining the two methods.
  • the first device may determine corresponding images to be recognized based on the recognition state and image quality, and then enter these images determined by these two methods as the final images to be recognized to step S13.
  • the first device can also determine the corresponding image to be recognized based on the recognition status and image quality, and then select a certain frame or frames as the final image to be recognized in a certain manner or randomly. Go to step S13.
  • the first device may also provide the user with two options of determining the image to be recognized according to the recognition state and determining the image to be recognized according to the image quality, that is, let the user choose which way to determine the image to be recognized.
  • the quality of the image to be recognized can be ensured, so that the first recognition result finally obtained based on the image to be recognized is more accurate and has higher accuracy.
  • selecting at least one frame of image as the image to be recognized may include: performing second information recognition on at least part of the image in the image sequence to obtain one or more first images. 2. Recognition results. According to one or more second recognition results, a reference recognition result is obtained. Each second recognition result is compared with the reference recognition result to obtain the confidence of each second recognition result. According to the confidence level, each image corresponding to at least a part of all the second recognition results is determined as the image to be recognized, wherein the determined image to be recognized has a higher confidence level than that of the unrecognized image in the image sequence. An image that is not to be recognized is an image that is not used for information recognition.
  • the second information recognition may be that the first device performs an information recognition operation on the image sequence of the object to be recognized to determine the information to be recognized.
  • the second device can obtain the first recognition result based on the information to be identified, indicating that the second device can perform the corresponding information recognition operation based on the information to be identified.
  • the operation performed by the second device according to the information to be recognized is recorded as the first information recognition, and the result obtained is recorded as the first recognition result, and the first recognition result is recorded as the first recognition result.
  • the operation performed by the device on at least part of the images in the image sequence is recorded as the second information recognition, and the obtained result is recorded as the second recognition result.
  • the specific recognition methods of the first information recognition and the second information recognition may be the same or different.
  • the second information recognition since the second information recognition is mainly to determine the image to be recognized rather than the final recognition result, the second information recognition may be different from the first information recognition, and the second information recognition may select the recognition accuracy.
  • the first information recognition can select the recognition method with higher recognition accuracy.
  • both the second information recognition and the first information recognition may be implemented by optical character recognition (OCR, Optical Character Recognition).
  • the scale of the OCR model used in the second information recognition may be smaller than the scale of the OCR model used in the first information recognition In this way, the realization of the second information recognition can be ensured, and the realization speed of the second information recognition can be ensured, and then the speed of the entire information recognition process can be improved.
  • the first device can perform second information recognition on each frame of image in the image sequence, or perform second information recognition on some images in the image sequence.
  • Which frame images are specifically selected for second information identification can be flexibly determined according to actual conditions, and there is no limitation in the embodiment of the present disclosure.
  • the subsequent disclosed embodiments all take the second information recognition of each frame of the image in the image sequence as an example for description, and other cases can be flexibly extended.
  • the second recognition result of each frame of the image can be obtained respectively.
  • the reference recognition result may be a relatively complete recognition result integrated based on at least two second recognition results.
  • some of the second recognition results may lack the first half of the content to be recognized, and some of the second recognition results lack the second half of the content to be recognized.
  • some second recognition results lack one or some fields of the content to be recognized. Therefore, when these second recognition results are counted, a relatively complete and accurate recognition result can be recovered, and this relatively complete and accurate recognition result can be used as a reference recognition result.
  • the specific recovery method is not limited in the embodiments of the present disclosure, and can be flexibly selected according to actual conditions.
  • the reference recognition result can be determined by traversing each second recognition result, counting the repetitive recognition content therein, and integrating according to the position of the repetitive recognition content.
  • each second recognition result can be compared with the reference recognition result to obtain the confidence level of each second recognition result.
  • the confidence degree represents the degree of overlap between each second recognition result and the reference recognition result.
  • the confidence level may represent the accuracy rate of the text overlap between each second recognition result and the reference recognition result. The higher the confidence level, the closer the second recognition result is to the reference recognition result.
  • the image to be recognized can be determined according to the degree of confidence. Since the higher the confidence level indicates that the second recognition result is closer to the reference recognition result, the image corresponding to the second recognition result with a higher confidence level contains more comprehensive recognition information, and the accuracy of information recognition based on this image is also Higher. It has been proposed in the above disclosed embodiments that the image to be recognized may be one frame or multiple frames, so one or more frames of images with the highest confidence can be used as the image to be recognized. Alternatively, a confidence threshold may be set, and the image corresponding to the second recognition result whose confidence is higher than the threshold is determined as the image to be recognized. Alternatively, the image corresponding to the second recognition result with the top 20% confidence level may be determined as the image to be recognized. These examples of determining the image to be recognized are not restrictive, and other methods or other values may also be used.
  • selecting at least one frame of image as the image to be recognized according to the image quality of the image sequence may include: separately acquiring image quality of at least part of the images in the image sequence in at least one measurement dimension. At least one frame of image in the image sequence whose image quality is greater than the threshold under the corresponding measurement dimension is determined as the image to be recognized.
  • the image quality of each frame of the image sequence can be obtained separately, or the image quality of only some frames in the image sequence can be obtained, which frames are specifically selected and how to choose, It can be flexibly determined according to actual conditions, and is not limited in the embodiments of the present disclosure.
  • the subsequent disclosed embodiments are described by taking the image quality of each frame of the image in the image sequence as an example, and the implementation of other situations can be flexibly extended with reference to the following disclosed embodiments.
  • the image quality can be the quality of the image in the image sequence.
  • the quality of the image can be judged by different judgment standards, such as clarity or completeness. Different angles to judge separately. Therefore, with different evaluation criteria, image quality can be analyzed under different measurement dimensions.
  • the image quality can be flexibly set according to the actual situation by analyzing the image quality through a certain measurement dimension or a certain number of measurement dimensions, and the specific evaluation standard corresponding to each measurement dimension.
  • the measurement dimensions may include one or more of: clarity, integrity, strong light conditions, dark light conditions, and occlusion conditions.
  • clarity, integrity, strong light conditions, dark light conditions, and occlusion conditions can indicate whether the image has focus blur, motion blur, etc. that cause the text or image to be unclear; the completeness can indicate whether the edges and corners of the object to be recognized (such as the document) in the image are all within the image range, etc.; Strong light conditions can indicate whether the image is overexposed or strongly reflective; dark light conditions can indicate whether the overall or partial brightness of the object to be identified in the image (such as a document) is too dark, causing the text or image to be unrecognizable; blocking conditions It can indicate whether the object to be identified in the image (such as a certificate) is blocked by other objects, etc.
  • the five measurement dimensions of clarity, integrity, strong light, dark light, and occlusion can be considered at the same time.
  • the image quality of each frame of image in the image sequence under these five measurement dimensions can be obtained separately, and then for each frame of image, it is considered whether its definition quality is greater than the corresponding definition threshold, and whether the integrity quality is Is greater than the corresponding integrity threshold, whether the strong light quality is greater than the corresponding strong light threshold, whether the dark light quality is greater than the corresponding dark light threshold, and whether the occlusion quality is greater than the corresponding occlusion threshold.
  • the image that meets the requirements can be used as the selected image.
  • the measurement dimensions include the above five dimensions
  • the image quality of some of the dimensions of the image can also be obtained. For example, only the three measurements of image clarity, completeness, and occlusion can be obtained. Quality in dimensions.
  • the subsequent determination of the image to be recognized may only consider whether the sharpness quality, integrity quality, and occlusion quality of each frame of the image are respectively greater than the corresponding threshold, and the comparison of the strong light quality and the dark light quality is omitted.
  • the image quality of these images that meet the requirements can be further compared, and N with a higher comprehensive image quality can be selected.
  • N is the required number of images to be recognized.
  • the calculation method of the comprehensive quality can be set according to the actual situation, and the specific calculation method may be selected according to the actual situation.
  • a weight can be set for the image quality under each measurement dimension, so as to calculate the weighted average image quality of each frame of image, and use the weighted average image quality as the comprehensive image quality.
  • the process of obtaining the image sequence of the object to be recognized can also be performed at the same time as the process of determining the selected image according to the image quality.
  • the image quality is judged.
  • the images that have been collected include images that meet the image quality requirements of each measurement dimension, such images can be used as the image to be recognized, and continue to collect the image sequence of the object to be recognized.
  • the last image in the image sequence that may be acquired is the image to be recognized.
  • the threshold value corresponding to the image quality in each measurement dimension is not limited in the embodiment of the present disclosure, and can be set according to actual conditions.
  • the threshold value of the image quality in different measurement dimensions may be the same or different.
  • At least one frame of image whose image quality in each measurement dimension is greater than the corresponding threshold is used as the image to be recognized.
  • the method proposed in the embodiment of the present disclosure may further include: saving at least one frame in the image to be recognized.
  • the storage location of the image to be recognized is not limited, and it can be stored in the first device, in the second device, or in the first device and the second device at the same time.
  • the saved image to be recognized can be directly read, which greatly improves the efficiency and user experience.
  • the number of saved frames of the image to be recognized is also not limited, and can be flexibly determined according to the size of the storage space of the first device and the second device.
  • all the images to be recognized can be saved, or one or several frames can be selected for saving; in a possible implementation, in When the image to be recognized is one frame, the image to be recognized can be directly saved.
  • the determination of the image to be recognized according to the recognition status and the determination of the image to be recognized according to the image quality can be implemented at the same time or separately, correspondingly, saving at least one frame of the image to be recognized can only be performed when the image to be recognized is determined according to the recognition status.
  • Post-realization that is, only save the to-be-recognized image determined according to the recognition state, or only after the to-be-recognized image is determined according to the image quality, that is, only save the to-be-recognized image determined by the After the quality of the image to be recognized is determined, it is possible to save the image to be recognized comprehensively determined according to the above two determination methods.
  • the image to be recognized determined based on the image quality has better image quality than the image to be recognized determined based on the recognition status, only the image determined by the image quality can be saved. The image to be recognized.
  • step S13 may be used to determine the information to be recognized based on the image to be recognized.
  • step S13 may include:
  • the implementation manners of the first encryption processing and the signature processing are not limited, and can be flexibly selected according to actual conditions.
  • the first encryption processing may include: encoding and encrypting the image to be recognized to obtain the first encrypted information; the signature processing may include: adding signature information of the first device to the selected image.
  • the encoding method used to encrypt the image to be recognized is not limited in the embodiment of the present disclosure, and any encryption method can be used as the implementation method of the first encryption process.
  • the signature information of the first device may be information containing the identity of the first device, and its specific information content and form are not limited in the embodiments of the present disclosure.
  • the signature information of the first device may be the signature information in a software development kit (SDK, Software Development Kit) in the first device.
  • SDK Software Development Kit
  • the specific location and manner of adding the signature information of the first device to the image to be recognized are not limited in the embodiment of the present disclosure, and can be selected flexibly according to the actual situation.
  • first encryption processing and the signature processing can be performed on the image to be recognized.
  • these two processings can be performed at the same time, or can be performed sequentially in a certain order, and can be flexibly selected according to actual conditions.
  • the first encryption processing and signature processing can be performed on the image to be identified in sequence. That is, the image to be identified is first re-encoded according to a certain encryption form, and then the information obtained by encoding and encryption is packaged with the signature information of the first device to obtain the information to be identified.
  • signature processing and first encryption processing on the image to be recognized in sequence. That is, the image to be recognized is packaged with the signature information of the first device first, and then the packaged information is re-encoded according to a certain encryption form to obtain the information to be recognized.
  • the object to be identified is an object with high security requirements such as an ID card or a bank card
  • the possibility of leakage of card information can be effectively reduced, thereby improving the security of the information identification process
  • the device information of the first device can be included in the information to be identified, so that the second device can verify the authorization of the information to be identified when performing information identification, and reduce the possibility of tampering with the information to be identified.
  • the security of the information recognition process is further improved, and the accuracy of the information recognition result is also improved.
  • step S14 may be used to send the information to be identified to the second device, and the specific sending mode is flexibly determined according to the connection relationship between the first device and the second device.
  • the first device when the first device is a terminal device and the second device is a server, the first device can upload to the second device through the network connected between the first device and the second device. Information to be identified.
  • the identification information fed back by the second device may also be received. Since the second device can obtain the first identification result according to the information to be identified, the identification information may be information related to the first identification result.
  • the identification information may be information related to the first identification result.
  • the manner in which the first device receives the identification information is not limited, and can also be flexibly determined according to the communication manner between the first device and the second device, which will not be repeated here.
  • the first device After receiving the identification information, the first device can display the identification information to the user, or save the identification information while displaying it.
  • the specific application of the identification information can be flexibly determined according to the actual needs of the first device. There is no limitation in the embodiment.
  • Fig. 2 shows a flowchart of an information identification method according to an embodiment of the present disclosure. The method can be applied to a second device.
  • the second device For the implementation of the second device, reference may be made to the above-mentioned disclosed embodiments, and details are not described herein again.
  • the information identification method may be applied to the second device, and includes: step S21, receiving information to be identified of the object to be identified.
  • step S22 Perform first information recognition on the to-be-identified information to obtain a first recognition result.
  • Step S23 Send identification information to the first device according to the first identification result.
  • the to-be-identified information is consistent with the to-be-identified information mentioned in the above disclosed embodiment, and will not be repeated here.
  • the manner in which the second device receives the information to be identified can be flexibly determined according to the manner in which the first device sends the information to be identified, which is not limited in the embodiment of the present disclosure.
  • the second device can perform the first information identification through step S22.
  • step S23 may be used to send the identification information to the first device according to the first recognition result, where the identification information may be obtained according to the first recognition result.
  • the first recognition result can be directly used as the identification information.
  • some other processing or content may be additionally performed according to the first recognition result to obtain the identification information.
  • the specific content and generation method of the identification information please refer to the subsequent disclosed embodiments.
  • the first information is recognized on the to-be-identified information to obtain the first recognition result, and the identification information is sent to the first device based on the first recognition result.
  • the information recognition of the object to be recognized is effectively realized. Since the first device realizes the frame selection function and the second device realizes the recognition function, both devices realize a single function and have lower requirements for calculation. Therefore, the cooperation of the two devices in the application can achieve faster processing speed and Stronger calculation effect, which greatly improves the efficiency and accuracy of the entire information recognition process.
  • step S22 may include: step S221, acquiring the image to be recognized included in the information to be recognized.
  • step S222 anti-counterfeiting detection is performed on the image to be identified, and the detection result is obtained.
  • Step S223 If the detection result is passed, perform first information recognition on the image to be recognized to obtain the first recognition result.
  • step S221 can be used to perform anti-counterfeiting detection on the image to be recognized.
  • the object to be identified is an object with high security requirements such as a card object
  • some users may use a copy or a duplicate of the object to be identified as the object to be identified for information identification.
  • the specific anti-counterfeiting detection method can be selected according to the actual situation, and refer to the subsequent disclosed embodiments. After the anti-counterfeiting detection is passed, the image to be identified can be identified by the first information. If the anti-counterfeiting detection fails, the detection failure or alarm notification can be fed back to the first device to further improve the security of the information recognition process.
  • the detection result is obtained, and if the detection result is passed, the image to be recognized is recognized by the first information to obtain the first recognition result.
  • step S221 can be flexibly selected according to actual conditions.
  • the information to be identified is an image to be identified
  • the image to be identified can be directly read from the information to be identified.
  • the to-be-identified information is information obtained by performing the first encryption processing on the to-be-identified image
  • the to-be-identified information can be decrypted to obtain the to-be-identified image.
  • step S221 may include: step S2211, acquiring signature information included in the information to be identified .
  • Step S2212 in the case where the acquired signature information matches the signature information of the first device, decrypt the first encrypted information included in the identification information to obtain the image to be identified.
  • the second device may first perform authorization verification on the identification information to be identified according to the signature information included in the identification information.
  • the signature information matches the signature information of the first device, it can indicate that the information to be identified has not been tampered with by other devices or users during the process of sending to the second device, that is, the image to be identified in the information to be identified It is an image that can be used for information recognition.
  • the second device can decrypt the first encrypted information, thereby restoring the image to be recognized determined by the first device.
  • the way of decryption can be flexibly determined according to the way of encryption, which is not limited in the embodiment of the present disclosure.
  • the signature information included in the to-be-identified information does not match the signature information of the first device, it means that the to-be-identified information may have been tampered with.
  • the equipment feeds back information such as matching failure or alarm prompts to further enhance the security of the information identification process.
  • the authorization verification of the information to be identified can be realized.
  • an alarm can be issued when the information to be identified is tampered with, on the other hand, it can reduce the meaningless decryption and identification process, and improve the security of information identification. And efficiency.
  • step S222 is not limited, that is, the manner of anti-counterfeiting detection of the image to be identified is not limited, and can be flexibly selected according to the actual situation.
  • it is possible to determine whether the image to be recognized is a frame image corresponding to the copy or the copy based on the unique characteristics of the copy or remake of the object to be recognized, for example, the remake may have reflections
  • the color of the copy may be significantly different from the original.
  • step S222 may include: step S2221, classifying the image to be recognized to obtain the classification result of the image to be recognized.
  • step S2222 when the classification result indicates that the image to be recognized is an image obtained by shooting the object to be recognized, the detection result is recorded as passed. And/or, in step S2223, when the classification result indicates that the image to be recognized is an image obtained by photographing a copy or a reproduction of the object to be recognized, the detection result is recorded as a failure.
  • the type of the image to be recognized can be determined by classifying the image to be recognized, so as to realize the anti-counterfeiting detection of the image to be recognized. How to classify the image to be recognized is not limited.
  • a classification neural network model may be used to classify the image to be recognized, wherein the specific implementation manner of the classification neural network model is not limited.
  • an initial neural network model can be established, and then a large number of images of the object to be identified, photocopies of the object to be identified, and duplicate images of the object to be identified obtained by shooting are used as training Sample, the initial neural network model is trained to obtain a trained classification neural network model.
  • a probability value can be output to indicate the probability that the image to be recognized is the original image of the object to be recognized.
  • this probability value is greater than the set probability threshold, it can indicate that the image to be recognized is the original image of the object to be recognized; otherwise, it indicates that the image to be recognized is a copy image or a duplicate image of the object to be recognized.
  • the probability threshold can be flexibly set according to actual conditions, and is not limited in the embodiments of the present disclosure.
  • the image to be recognized is classified to obtain the classification result, and when the classification result indicates that the image to be recognized is an image obtained by shooting the object to be recognized, the detection result is recorded as a pass, otherwise it is recorded as a failure.
  • the classification method can be used to realize the anti-counterfeiting detection of the image to be recognized, which has both high detection efficiency and high detection accuracy, thereby greatly improving the accuracy and speed of the entire information recognition process.
  • step S223 may include: performing OCR recognition on the image to be recognized through the optical character recognition OCR model to obtain the first recognition result.
  • the purpose of the first information recognition is to recognize the corresponding information contained in the image to be recognized, it has a higher recognition accuracy requirement. Therefore, in the embodiment of the present disclosure, a larger The large-scale OCR model realizes the first information recognition and improves the accuracy and accuracy of information recognition.
  • step S22 may further include: step S224, judging whether the first recognition result matches a preset rule, and obtaining the judgment result.
  • the preset rules may be certain verification rules determined according to the characteristics of the information in the object to be identified.
  • this encoding rule can be used as a preset rule .
  • the coding rule of the bank card number can also be changed as a preset rule.
  • the judgment result can be recorded as a match passed, and when the first recognition result does not meet the preset rule, the judgment result can be recorded as a match failed.
  • the first recognition result can be further verified, so that when the first recognition result does not match the preset rule, a certain prompt or warning is issued so that the user can confirm whether the recognition result is accurate and whether it needs to be renewed. Recognition etc.
  • step S23 After the first identification result is obtained, the identification information can be sent to the first device through step S23 based on the first identification result.
  • the implementation of step S23 can be flexibly determined according to the actual situation.
  • the step S23 may include: step S231, performing second encryption processing on the first recognition result, or performing second encryption processing on the first recognition result and the judgment result, to obtain second encrypted information; step S232, using the second encrypted information as the recognition Information and send to the first device.
  • the second encryption process may be directly performed on the first identification result to obtain the second encrypted information, so that the second encrypted information is sent to the first device as the identification information.
  • the "first" and "second" in the second encryption process and the first encryption process are only used to distinguish the subject and object of the encryption process, that is, the first encryption process is the encryption of the image to be recognized by the first device.
  • the second encryption process is the encryption of the first recognition result by the second device, and does not limit whether the encryption methods are the same, that is, the encryption rules of the first encryption process and the second encryption process can be the same or different, and can be selected flexibly according to the actual situation. That's it.
  • the judgment result can be packaged with the first recognition result, and the packaged information can be carried out together.
  • the second encryption process is performed to obtain the second encrypted information, and the second encrypted information is sent to the first device as identification information, so that the user or the first device can make a decision whether or not to perform information identification again according to the judgment result.
  • the information recognition method applied to the second device proposed in the embodiment of the present disclosure may further include: saving at least one frame in the image to be recognized.
  • the second device can recover the to-be-recognized information from the received information to be recognized.
  • the image is recognized. Therefore, the second device can also save at least one frame in the image to be recognized.
  • the second device when the second device saves the image to be recognized, whether it specifically saves one frame of the image to be recognized or multiple frames of the image to be recognized can be flexibly determined according to the actual situation, and will not be repeated here. Since the image to be recognized may be the image selected by the first device based on the recognition state, or it may be the image selected by the first device based on image quality in multiple measurement dimensions. With the different selection methods, the image to be recognized The quality can also change accordingly. Therefore, in a possible implementation, the second device can choose different storage methods according to the different selection methods of the image to be recognized. The specific selection can be carried out according to the actual situation. Flexible settings are not limited to the following disclosed embodiments.
  • the second device may choose to save only the image quality determined based on multiple measurement dimensions. To improve the quality of the to-be-recognized image saved in the second device.
  • the multiple images to be recognized can also be subjected to another image quality screening, so that the image with the highest image quality is selected. It is stored in the second device, and the method of image quality screening in the second device can be the same as or different from that in the first device, and it can be flexibly selected according to the actual situation.
  • the image to be recognized is retrieved from the device to reduce user operations and improve user experience.
  • FIG. 3 shows a block diagram of an information recognition device 30 according to an embodiment of the present disclosure.
  • the device 30 may include: an image sequence acquiring module 31, configured to acquire an image sequence of an object to be identified, the image sequence including at least two frames of images.
  • the determining module 32 is configured to determine at least one frame of image from the image sequence as the image to be recognized.
  • the to-be-identified information generating module 33 is configured to determine the to-be-identified information according to the image to be identified.
  • the to-be-identified information sending module 34 is configured to send the to-be-identified information to the second device so that the second device can obtain the first identification result according to the to-be-identified information.
  • the determining module is used to determine at least one frame of image from the image sequence as the image to be recognized according to the recognition state and/or image quality of the image sequence.
  • the determining module is further used to: perform second information recognition on at least part of the images in the image sequence to obtain one or more second recognition results; and obtain one or more second recognition results according to the one or more second recognition results.
  • Reference recognition results respectively compare each second recognition result with the reference recognition result to obtain the confidence level of each second recognition result; according to the confidence level, determine each image corresponding to at least a part of all the second recognition results as The image to be recognized, wherein the determined confidence of the image to be recognized is higher than the confidence of the image that is not to be recognized in the image sequence.
  • the determining module is further configured to: obtain the image quality of at least part of the images in the image sequence in at least one measurement dimension; The image is determined to be the image to be recognized.
  • the measurement dimension includes at least one of sharpness, completeness, strong light conditions, dark light conditions, and occlusion conditions.
  • the device 30 further includes a first saving module, and the first saving module is configured to save at least one frame in the image to be recognized.
  • the to-be-identified information generating module is used to perform first encryption processing and/or signature processing on the to-be-identified image to obtain the to-be-identified information.
  • the first encryption processing includes: encoding and encrypting the image to be recognized to obtain the first encrypted information; the signature processing includes: adding signature information of the first device to the image to be recognized.
  • the object to be recognized includes a card object and/or a form object; the first recognition result includes at least one of text, characters, logos, icons, and pictures recorded in the object to be recognized.
  • Fig. 4 shows a block diagram of an information recognition device according to an embodiment of the present disclosure.
  • the device 40 may include: a receiving module 41, configured to receive the to-be-identified information of the to-be-identified object.
  • the identification module 42 is configured to perform first information identification on the information to be identified to obtain a first identification result.
  • the identification information sending module 43 is configured to send identification information to the first device according to the first identification result.
  • the recognition module is used to: obtain the image to be identified included in the information to be identified; perform anti-counterfeiting detection on the image to be identified to obtain the detection result; if the detection result is passed, perform the first step on the image to be identified The information is recognized, and the first recognition result is obtained.
  • the identification module is further configured to: obtain the signature information included in the information to be identified; in the case that the obtained signature information matches the signature information of the first device, the first encrypted information included in the identification information is Decryption is performed to obtain the image to be recognized.
  • the recognition module is further used to: classify the image to be recognized to obtain a classification result of the image to be recognized; in the case where the classification result indicates that the image to be recognized is an image obtained by shooting the object to be recognized, Record the detection result as a pass; and/or record the detection result as a failure when the classification result indicates that the image to be recognized is an image obtained by shooting a photocopy or remake of the object to be recognized.
  • the recognition module is further used to perform OCR recognition on the image to be recognized through the optical character recognition OCR model to obtain the first recognition result.
  • the recognition module is further used to determine whether the first recognition result matches a preset rule, and obtain the determination result.
  • the identification information sending module is used to: perform a second encryption process on the first recognition result, or perform the second encryption process on the first recognition result and the judgment result, to obtain the second encrypted information ; Use the second encrypted information as the identification information and send it to the first device.
  • the device 40 further includes a second saving module, and the second saving module is configured to save at least one frame in the image to be recognized.
  • the embodiment of the present disclosure also discloses an application example.
  • the application example proposes an information recognition system. Based on this information recognition system, business processes such as online loans, renting, or membership registration can be used. , To realize the identification and collection of ID card information.
  • FIG. 5 shows a schematic diagram of an application example according to the present disclosure.
  • the information recognition system is mainly composed of a first device at the front end and a second device at the back end, wherein the first device at the front end (hereinafter referred to as the front end) It can be a mobile phone device, a tablet device, a notebook or computer device with a camera, etc., and the second back-end device (hereinafter referred to as the back-end) can be a cloud server.
  • the first device can obtain the image sequence of the object to be identified, and select at least one frame from the image sequence as the to-be-identified object. Recognize the image, and then determine the to-be-identified information based on the to-be-identified image to send the to-be-identified information to the second device. After receiving the to-be-identified information, the second device can perform the first information recognition on the to-be-identified information to obtain the first recognition result, And send identification information to the first device according to the first identification result.
  • the process of identifying and collecting ID card information by the information identification system may be:
  • the front end can turn on the camera to continuously scan and select the frame of the document to be recognized until a frame that meets the conditions is selected as the image to be recognized, and then the selected image to be recognized is transmitted to the back end for back-end processing.
  • the image to be identified can be encoded and encrypted, and the signature information in the front-end SDK is attached at the same time for permission verification to ensure that the image to be identified is not tampered with during the data transmission process.
  • the back end After receiving the image to be recognized from the front end, the back end can perform decryption and permission verification, and after decryption and permission verification confirm that the data has not been tampered with, it enters the recognition process.
  • the specific identification process can be as follows: firstly, perform anti-counterfeiting detection on the document image through the neural network model to determine whether the document is the original; secondly, perform text recognition on the ID card, and use OCR technology to identify the various fields of the document; and finally according to the preset Logic, verify and judge the results of character recognition. After the recognition is completed, the results of the text recognition and the logical judgment are encrypted and transmitted back to the front end and provided to the user.
  • the process of selecting frames by the front end can be implemented in one of the following two ways.
  • Method 1 Select the frame based on the text recognition result of each frame of the collected image sequence by the front end
  • the front-end OCR small model can be used to perform text recognition on each frame of image collected by the camera, and the quality of the frame can be judged according to the confidence of the text recognition result.
  • the front end performs text recognition on each frame of image to obtain multiple text recognition results, and then integrates the multiple text recognition results into a standard recognition result to filter multiple frames of images, that is, select the text recognition result that is closest to the standard
  • An image of the recognition result is used as the image to be recognized and is handed over to the back-end for processing. Due to the limited accuracy of the OCR small model and the limited size of the font, the image to be recognized still needs to be transmitted to the back end for the recognition of the OCR large model. Through this method of selecting frames, it can have better tolerance, and can have better frame selection results under different quality cameras and light environments.
  • This type of frame selection can perform the quality inspection of the document image from the five dimensions of clarity, integrity, strong light, dark light, and occlusion.
  • the definitions of these five dimensions are:
  • Sharpness mainly describes whether the image has focus blur, motion blur, etc. that cause unclear text or image recognition; completeness, mainly describes whether the corners of the document are all within the image range; strong light conditions, mainly describes whether the document exists Overexposure or strong reflection; dim light situation, mainly describes whether the overall or partial brightness of the document is too dark, which causes the text or image to be unrecognizable; occlusion situation, mainly describes whether the document in the image is blocked by other objects.
  • each frame of image can be detected in the above five dimensions, and the detection result can be compared with a preset threshold (the preset threshold can be adjusted according to the actual situation).
  • the frame selection result can be output for subsequent back-end recognition.
  • the writing order of the steps does not mean a strict execution order but constitutes any limitation on the implementation process.
  • the specific execution order of each step should be based on its function and possibility.
  • the inner logic is determined.
  • the embodiments of the present disclosure also provide a computer-readable storage medium on which computer program instructions are stored, and the computer program instructions implement the above-mentioned method when executed by a processor.
  • the computer-readable storage medium may be a volatile computer-readable storage medium or a non-volatile computer-readable storage medium.
  • An embodiment of the present disclosure also provides an electronic device, including: a processor; a memory for storing executable instructions of the processor; wherein the processor is configured as the above-mentioned method.
  • the above-mentioned memory may be a volatile memory (volatile memory), such as RAM; or a non-volatile memory (non-volatile memory), such as ROM, flash memory, or hard disk (Hard Disk Drive). , HDD) or solid-state drive (Solid-State Drive, SSD); or a combination of the above types of memory, and provide instructions and data to the processor.
  • volatile memory such as RAM
  • non-volatile memory such as ROM, flash memory, or hard disk (Hard Disk Drive). , HDD) or solid-state drive (Solid-State Drive, SSD); or a combination of the above types of memory, and provide instructions and data to the processor.
  • the foregoing processor may be at least one of ASIC, DSP, DSPD, PLD, FPGA, CPU, controller, microcontroller, and microprocessor. It is understandable that, for different devices, the electronic device used to implement the above-mentioned processor function may also be other, and the embodiment of the present disclosure does not specifically limit it.
  • the electronic device can be provided as a terminal, server or other form of device.
  • the embodiment of the present disclosure also provides a computer program, which implements the foregoing method when the computer program is executed by a processor.
  • FIG. 6 is a block diagram of an electronic device 800 according to an embodiment of the present disclosure.
  • the electronic device 800 may be a mobile phone, a computer, a digital broadcasting terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, and other terminals.
  • the electronic device 800 may include one or more of the following components: a processing component 802, a memory 804, a power supply component 806, a multimedia component 808, an audio component 810, an input/output (I/O) interface 812, a sensor component 814, And the communication component 816.
  • the processing component 802 generally controls the overall operations of the electronic device 800, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations.
  • the processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the foregoing method.
  • the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components.
  • the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.
  • the memory 804 is configured to store various types of data to support operations in the electronic device 800. Examples of these data include instructions for any application or method to operate on the electronic device 800, contact data, phone book data, messages, pictures, videos, etc.
  • the memory 804 can be implemented by any type of volatile or non-volatile storage device or their combination, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic Disk or Optical Disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read-only memory
  • EPROM erasable Programmable Read Only Memory
  • PROM Programmable Read Only Memory
  • ROM Read Only Memory
  • Magnetic Memory Flash Memory
  • Magnetic Disk Magnetic Disk or Optical Disk.
  • the power supply component 806 provides power for various components of the electronic device 800.
  • the power supply component 806 may include a power management system, one or more power supplies, and other components associated with the generation, management, and distribution of power for the electronic device 800.
  • the multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user.
  • the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user.
  • the touch panel includes one or more touch sensors to sense touch, sliding, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure related to the touch or slide operation.
  • the multimedia component 808 includes a front camera and/or a rear camera. When the electronic device 800 is in an operation mode, such as a shooting mode or a video mode, the front camera and/or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have focal length and optical zoom capabilities.
  • the audio component 810 is configured to output and/or input audio signals.
  • the audio component 810 includes a microphone (MIC), and when the electronic device 800 is in an operation mode, such as a call mode, a recording mode, and a voice recognition mode, the microphone is configured to receive an external audio signal.
  • the received audio signal may be further stored in the memory 804 or transmitted via the communication component 816.
  • the audio component 810 further includes a speaker for outputting audio signals.
  • the I/O interface 812 provides an interface between the processing component 802 and a peripheral interface module.
  • the above-mentioned peripheral interface module may be a keyboard, a click wheel, a button, and the like. These buttons may include, but are not limited to: home button, volume button, start button, and lock button.
  • the sensor component 814 includes one or more sensors for providing the electronic device 800 with various aspects of state evaluation.
  • the sensor component 814 can detect the on/off status of the electronic device 800 and the relative positioning of the components.
  • the component is the display and the keypad of the electronic device 800.
  • the sensor component 814 can also detect the electronic device 800 or the electronic device 800.
  • the position of the component changes, the presence or absence of contact between the user and the electronic device 800, the orientation or acceleration/deceleration of the electronic device 800, and the temperature change of the electronic device 800.
  • the sensor component 814 may include a proximity sensor configured to detect the presence of nearby objects when there is no physical contact.
  • the sensor component 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications.
  • the sensor component 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
  • the communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices.
  • the electronic device 800 can access a wireless network based on a communication standard, such as WiFi, 2G, or 3G, or a combination thereof.
  • the communication component 816 receives a broadcast signal or broadcast related personnel information from an external broadcast management system via a broadcast channel.
  • the communication component 816 further includes a near field communication (NFC) module to facilitate short-range communication.
  • the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID radio frequency identification
  • IrDA infrared data association
  • UWB ultra-wideband
  • Bluetooth Bluetooth
  • the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASIC), digital signal processors (DSP), digital signal processing devices (DSPD), programmable logic devices (PLD), field-available A programmable gate array (FPGA), controller, microcontroller, microprocessor, or other electronic components are implemented to implement the above methods.
  • ASIC application-specific integrated circuits
  • DSP digital signal processors
  • DSPD digital signal processing devices
  • PLD programmable logic devices
  • FPGA field-available A programmable gate array
  • controller microcontroller, microprocessor, or other electronic components are implemented to implement the above methods.
  • a non-volatile computer-readable storage medium such as the memory 804 including computer program instructions, which can be executed by the processor 820 of the electronic device 800 to complete the foregoing method.
  • FIG. 7 is a block diagram of an electronic device 1900 according to an embodiment of the present disclosure.
  • the electronic device 1900 may be provided as a server. 7
  • the electronic device 1900 includes a processing component 1922, which further includes one or more processors, and a memory resource represented by a memory 1932, for storing instructions that can be executed by the processing component 1922, such as application programs.
  • the application program stored in the memory 1932 may include one or more modules each corresponding to a set of instructions.
  • the processing component 1922 is configured to execute instructions to perform the above-described methods.
  • the electronic device 1900 may also include a power supply component 1926 configured to perform power management of the electronic device 1900, a wired or wireless network interface 1950 configured to connect the electronic device 1900 to a network, and an input output (I/O) interface 1958 .
  • the electronic device 1900 may operate based on an operating system stored in the memory 1932, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM or the like.
  • a non-volatile computer-readable storage medium is also provided, such as the memory 1932 including computer program instructions, which can be executed by the processing component 1922 of the electronic device 1900 to complete the foregoing method.
  • the present disclosure provides a system, method and/or computer program product.
  • the computer program product may include a computer-readable storage medium loaded with computer-readable program instructions for enabling a processor to implement various aspects of the present disclosure.
  • the computer-readable storage medium may be a tangible device that can hold and store instructions used by the instruction execution device.
  • the computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing.
  • Computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) Or flash memory), static random access memory (SRAM), portable compact disk read-only memory (CD-ROM), digital versatile disk (DVD), memory stick, floppy disk, mechanical encoding device, such as a printer with instructions stored thereon
  • RAM random access memory
  • ROM read-only memory
  • EPROM erasable programmable read-only memory
  • flash memory flash memory
  • SRAM static random access memory
  • CD-ROM compact disk read-only memory
  • DVD digital versatile disk
  • memory stick floppy disk
  • mechanical encoding device such as a printer with instructions stored thereon
  • the computer-readable storage medium used here is not interpreted as the instantaneous signal itself, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (for example, light pulses through fiber optic cables), or through wires Transmission of electrical signals.
  • the computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to various computing/processing devices, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and/or a wireless network.
  • the network may include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers.
  • the network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network, and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing/processing device .
  • the computer program instructions used to perform the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or in one or more programming languages.
  • Source code or object code written in any combination, the programming language includes object-oriented programming languages such as Smalltalk, C++, etc., and conventional procedural programming languages such as "C" language or similar programming languages.
  • the computer-readable program instructions can be executed entirely on the user's computer, partly on the user's computer, executed as a stand-alone software package, partly on the user's computer and partly executed on a remote computer, or entirely on the remote computer or server implement.
  • the remote computer can be connected to the user's computer through any kind of network-including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect to the user's computer) connect).
  • LAN local area network
  • WAN wide area network
  • an electronic circuit such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), is personalized by using status personnel information of computer-readable program instructions.
  • FPGA field programmable gate array
  • PDA programmable logic array
  • the computer-readable program instructions can be executed to implement various aspects of the present disclosure.
  • These computer-readable program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine that makes these instructions when executed by the processor of the computer or other programmable data processing device , A device that implements the functions/actions specified in one or more blocks in the flowcharts and/or block diagrams is produced. It is also possible to store these computer-readable program instructions in a computer-readable storage medium. These instructions make computers, programmable data processing apparatuses, and/or other devices work in a specific manner, so that the computer-readable medium storing the instructions includes An article of manufacture, which includes instructions for implementing various aspects of the functions/actions specified in one or more blocks in the flowcharts and/or block diagrams.
  • each block in the flowchart or block diagram may represent a module, program segment, or part of an instruction, and the module, program segment, or part of an instruction contains one or more components for realizing the specified logical function.
  • Executable instructions may also occur in a different order from the order marked in the drawings. For example, two consecutive blocks can actually be executed substantially in parallel, or they can sometimes be executed in the reverse order, depending on the functions involved.
  • each block in the block diagram and/or flowchart, and the combination of the blocks in the block diagram and/or flowchart can be implemented by a dedicated hardware-based system that performs the specified functions or actions Or it can be realized by a combination of dedicated hardware and computer instructions.

Abstract

本公开涉及一种信息识别方法及装置、系统、电子设备和存储介质。所述信息识别方法包括：第一设备获取待识别对象的图像序列，所述图像序列包括至少两帧图像；从所述图像序列中确定至少一帧图像，作为待识别图像；根据所述待识别图像，确定待识别信息；向第二设备发送所述待识别信息，以使所述第二设备根据所述待识别信息得到第一识别结果。所述信息识别方法还包括：第二设备接收待识别对象的待识别信息；根据所述待识别信息，进行第一信息识别，得到第一识别结果；根据所述第一识别结果，向第一设备发送识别信息。

Description

信息识别方法及装置、系统、电子设备、存储介质和计算机程序
相关申请的交叉引用
本申请要求2020年2月28日提交的中国专利申请202010129842.2的优先权，该中国专利申请的全部内容通过引用的形式并入本文。
技术领域
本公开涉及计算机视觉领域，尤其涉及一种信息识别方法及装置、系统、电子设备、存储介质和计算机程序。
背景技术
随着终端设备的不断发展，越来越多的业务可以在线上远程完成。由于很多业务需要验证用户的身份或是录入用户的证件信息，如果由用户手动输入证件信息或上传证件照片，交互流程体验较差，出错率高，证件信息也容易被篡改为安全性要求较高的业务留下了隐患。
因此，为了提高用户体验，确保业务的安全性，需要实现更加高效的信息录入。
发明内容
本公开提出了一种信息识别方案。
根据本公开的第一方面，提供了一种信息识别方法，包括：第一设备获取待识别对象的图像序列，所述图像序列包括至少两帧图像；所述第一设备从所述图像序列中确定至少一帧图像，作为待识别图像；根据所述待识别图像，确定待识别信息；向第二设备发送所述待识别信息，以使所述第二设备根据所述待识别信息得到第一识别结果。
在一种可能的实现方式中，所述第一设备从所述图像序列中确定至少一帧图像，作为待识别图像，包括：所述第一设备根据所述图像序列的识别状态和/或图像质量，从图像序列中确定至少一帧图像，作为待识别图像。
在一种可能的实现方式中，所述第一设备根据所述图像序列的识别状态，从所述图像序列中选定至少一帧图像，作为所述待识别图像，包括：第一设备对所述图像序列中至少部分图像分别进行第二信息识别，得到一个或多个第二识别结果；第一设备根据所述一个或多个第二识别结果，得到参考识别结果；分别将每个所述第二识别结果与所述参考识别结果进行比较，得到每个所述第二识别结果的置信度；第一设备根据所述置信度，将全部第二识别结果的至少一部分对应的各图像，确定为所述待识别图像，其中，确定出的所述待识别图像的置信度高于所述图像序列中的非待识别图像的置信度。
在一种可能的实现方式中，所述第一设备根据所述图像序列的图像质量，从所述图像序列中选定至少一帧图像，作为待识别图像，包括：第一设备分别获取所述图像序列中至少部分图像在至少一个衡量维度下的图像质量；第一设备将所述图像序列中所述图像质量大于对应衡量维度下的阈值的至少一帧图像，确定作为所述待识别图像。
在一种可能的实现方式中，所述衡量维度包括清晰度、完整度、强光情况、暗光情况以及遮挡情况中的至少一项。
在一种可能的实现方式中，在确定所述待识别图像之后，所述方法还包括：保存所述待识别图像中的至少一帧。
在一种可能的实现方式中，所述第一设备根据所述待识别图像，确定待识别信息，包括：所述第一设备对所述待识别图像进行第一加密处理和/或签名处理，得到待识别信息。
在一种可能的实现方式中，所述第一加密处理包括：对所述待识别图像进行编码加密，得到第一加密信息；所述签名处理包括：将所述第一设备的签名信息添加至所述待 识别图像。
在一种可能的实现方式中，所述待识别对象包括证卡对象和/或表单对象；所述第一识别结果包括所述待识别对象中记录的文本、标识以及图片中的至少一项。
根据本公开的第二方面，提供了一种信息识别方法，包括：第二设备接收待识别对象的待识别信息；所述第二设备对所述待识别信息进行第一信息识别，得到第一识别结果；所述第二设备根据所述第一识别结果，向第一设备发送识别信息。
在一种可能的实现方式中，所述第二设备对所述待识别信息进行第一信息识别，得到第一识别结果，包括：所述第二设备获取所述待识别信息包括的待识别图像；对所述待识别图像进行防伪检测，得到检测结果；在所述检测结果为通过的情况下，所述第二设备对所述待识别图像进行第一信息识别，得到第一识别结果。
在一种可能的实现方式中，所述第二设备获取所述待识别信息包括的待识别图像，包括：所述第二设备获取所述待识别信息包括的签名信息；在获取的所述签名信息与所述第一设备的签名信息匹配的情况下，所述第二设备对所述待识别信息包括的第一加密信息进行解密，得到所述待识别图像。
在一种可能的实现方式中，所述第二设备对所述待识别图像进行防伪检测，得到检测结果，包括：所述第二设备对所述待识别图像进行分类，得到所述待识别图像的分类结果；在所述分类结果指示所述待识别图像为通过对所述待识别对象进行拍摄得到的图像的情况下，所述第二设备将检测结果记录为通过；和/或，在所述分类结果指示所述待识别图像为通过对所述待识别对象的复印件或翻拍件进行拍摄得到的图像的情况下，所述第二设备将检测结果记录为失败。
在一种可能的实现方式中，所述第二设备对所述待识别图像进行第一信息识别，得到第一识别结果，包括：所述第二设备通过光学字符识别OCR模型，对所述待识别图像进行OCR识别，得到第一识别结果。
在一种可能的实现方式中，所述第二设备对所述待识别信息进行第一信息识别，得到第一识别结果，还包括：所述第二设备判断所述第一识别结果是否与预设规则匹配，得到判断结果。
在一种可能的实现方式中，所述第二设备根据所述第一识别结果，向第一设备发送识别信息，包括：所述第二设备对所述第一识别结果进行第二加密处理，或是对所述第一识别结果和所述判断结果进行所述第二加密处理，得到第二加密信息；所述第二设备将所述第二加密信息作为所述识别信息，并向所述第一设备发送。
在一种可能的实现方式中，所述方法还包括：所述第二设备保存所述待识别图像中的至少一帧。
根据本公开的第三方面，提供了一种信息识别装置，包括：图像序列获取模块，用于获取待识别对象的图像序列，所述图像序列包括至少两帧图像；确定模块，用于从所述图像序列中确定至少一帧图像，作为待识别图像；待识别信息生成模块，用于根据所述待识别图像，确定待识别信息；待识别信息发送模块，用于向第二设备发送所述待识别信息，以使所述第二设备根据所述待识别信息得到第一识别结果。
根据本公开的第四方面，提供了一种信息识别装置，包括：接收模块，用于接收待识别对象的待识别信息；识别模块，用于对所述待识别信息进行第一信息识别，得到第一识别结果；识别信息发送模块，用于根据所述第一识别结果，向第一设备发送发送识别信息。
根据本公开的第五方面，提供了一种电子设备，包括：处理器；用于存储处理器可 执行指令的非暂时性存储介质；其中，所述处理器被配置为：执行上述第一方面的信息识别方法。
根据本公开的第六方面，提供了一种电子设备，包括：处理器；用于存储处理器可执行指令的非暂时性存储介质；其中，所述处理器被配置为：执行上述第二方面的信息识别方法。
根据本公开的第七方面，提供了一种计算机可读存储介质，其上存储有计算机程序指令，所述计算机程序指令被处理器执行时实现上述第一方面的信息识别方法。
根据本公开的第八方面，提供了一种计算机可读存储介质，其上存储有计算机程序指令，所述计算机程序指令被处理器执行时实现上述第二方面的信息识别方法。
根据本公开的第九方面，提供了一种计算机程序，该计算机程序被处理器执行时实现以上任一方面的信息识别方法。
在本公开实施例中，通过获取待识别对象的图像序列，并根据图像序列选定至少一帧图像作为待识别图像，从而根据待识别图像向第二设备发送待识别信息，以使第二设备根据待识别信息来得到第一识别结果。通过上述过程，可以在信息识别的过程中，首先对待识别对象自动选帧，再通过第二设备进行识别。由于自动选帧可以使得用于识别的图像比起直接获取的图像具有更高的图像精度和识别效果，因此可以使得最终得到的第一识别结果更为准确，同时也可以降低信息识别过程的失败率，从而提升信息识别过程中的用户体验。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，而非限制本公开。根据下面参考附图对示例性实施例的详细说明，本公开的其它特征及方面将变得清楚。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，这些附图示出了符合本公开的实施例，并与说明书一起用于说明本公开的技术方案。
图1示出根据本公开一实施例的信息识别方法的流程图。
图2示出根据本公开一实施例的信息识别方法的流程图。
图3示出根据本公开一实施例的信息识别装置的框图。
图4示出根据本公开一实施例的信息识别装置的框图。
图5示出根据本公开一应用示例的示意图。
图6示出根据本公开实施例的一种电子设备的框图。
图7示出根据本公开实施例的一种电子设备的框图。
具体实施方式
以下将参考附图详细说明本公开的各种示例性实施例、特征和方面。附图中相同的附图标记表示功能相同或相似的元件。尽管在附图中示出了实施例的各种方面，但是除非特别指出，不必按比例绘制附图。
在这里专用的词“示例性”意为“用作例子、实施例或说明性”。这里作为“示例性”所说明的任何实施例不必解释为优于或好于其它实施例。
本文中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中术语“至少一种”表示多种中的任意一种或多种中的至少两种的任意组合，例如，包括A、B、C中的至少一种，可以表示包括从A、B和C构成的集合中选择的任意一个或多个元素。
另外，为了更好地说明本公开，在下文的具体实施方式中给出了众多的具体细节。本领域技术人员应当理解，没有某些具体细节，本公开同样可以实施。在一些实例中，对于本领域技术人员熟知的方法、手段、元件和电路未作详细描述，以便于凸显本公开的主旨。
图1示出根据本公开一实施例的信息识别方法的流程图，该方法可以应用于第一设备，第一设备可以是能够采集待识别对象的图像序列的设备，即具有信息采集功能的设备。在一种可能的实现方式中，第一设备可以是具备拍摄功能的终端设备或者其他处理设备等硬件设备。其中，终端设备可以为用户设备(User Equipment，UE)、移动设备、用户终端、终端、蜂窝电话、无绳电话、个人数字处理(Personal Digital Assistant，PDA)、手持设备、计算设备、车载设备、可穿戴设备等。
在一些可能的实现方式中，采集待识别对象的图像序列的操作也可以由诸如摄像头等的单独的前端设备来执行，然后前端设备将采集到的待识别对象的图像序列发送给第一设备。接着，通过第一设备的处理器调用存储器中存储的计算机可读指令的方式来实现该信息识别方法。
如图1所示，在一种可能的实现方式中，所述信息识别方法可以应用于第一设备，包括以下步骤。
步骤S11，获取待识别对象的图像序列，其中，图像序列包括至少两帧图像。
步骤S12，从图像序列中确定至少一帧图像，作为待识别图像。
步骤S13，根据所述待识别图像，确定待识别信息。
步骤S14，向第二设备发送待识别信息，以使第二设备根据待识别信息得到第一识别结果。
其中，待识别对象是要进行信息识别的对象，根据实际的信息识别需求，待识别对象可以具有多种表现形式。例如，待识别对象可以是诸如证卡对象或表单对象等内容具有统一规格的对象。证卡对象可以是身份证、通行证、银行卡等，表单对象可以是规定制式的表格等。
待识别对象的图像序列可以是第一设备对待识别对象进行图像采集所得到的图像序列或一组图像。在一种可能的实现方式中，第一设备可以通过对待识别对象进行持续的扫描或是视频录制，来获取待识别对象的图像序列。在一种可能的实现方式中，第一设备也可以通过对待识别对象按照一定的频率进行拍照采集，来获取待识别对象的多帧图像。所采集的多帧图像可以按照采集时间的先后顺序组成图像序列。当然，针对待识别对象采集的多帧图像彼此也可以不具有时间的关联性，此时，可以构成一组图像而非一个图像序列。
在步骤S12中，待识别图像可以是第一设备从获取的图像序列中选定的一帧或多帧图像。在一种可能的实现方式中，可以从图像序列中仅仅选定一帧具有较高质量的图像作为待识别图像。在一种可能的实现方式中，可以从图像序列中选定质量较高的多帧图像作为待识别图像。所选择的待处理图像的数量可以根据实际情况进行灵活确定。后续，以选定一帧图像为待识别图像的情况为例进行说明。对于选定多帧图像为待识别图像的情况，其实现原理与选定一帧图像的情况相类似，不再进行详细说明。然而，应当理解，选定多帧图像为待识别图像的情况也被包含在本公开的记载范围之内。
在步骤S13中，在确定了待识别图像后，可以基于待识别图像来确定待识别信息。待识别信息可以是包含有待识别对象中需要被识别内容的信息。在一种可能的实现方式中，待识别信息可以直接为选定的待识别图像。在一种可能的实现方式中，待识别信息 也可以是对待识别图像进行了一定的处理后所得到的信息。待识别信息的确定方式可以参见后续各公开实施例。
在步骤S14中，可以向第二设备发送待识别信息，以使第二设备根据待识别信息来得到第一识别结果。其中，第二设备可以是具有信息识别功能的设备。第二设备可以通过硬件或软件来实现。在一种可能的实现方式中，第二设备可以是诸如终端设备、服务器或者其他处理设备等的硬件设备。其中，终端设备的实现方式可以参考上述各公开实施例，在此不再赘述。在一种可能的实现方式中，当第二设备为服务器的情况下，第二设备可以是云端服务器，也可以是本地服务器等。在第二设备通过软件实现的情况下，第二设备的功能可以通过诸如CPU的处理器读取存储介质中的计算机程序来实现。
第一识别结果可以是第二设备根据待识别信息进行识别所获得的结果，具体的第一识别结果的种类和内容可以根据待识别对象的表现方式而不同，在本公开实施例中不做限制。在一种可能的实现方式中，待识别对象可以包括证卡对象和/或表单对象，第一识别结果可以包括待识别对象中记录的文本、字符、图标、标识以及图片等中的至少一项。由于诸如证卡对象或表单对象的待识别对象具有一定格式规范，因此可以基于规范化的内容完成第一识别。其中，证卡对象可以包括但不限于身份证、银行卡、通行证；表单对象可以包括但不限于保单、发票。得到的第一识别结果往往可以基于实际需求，以及相应的待识别对象所具备的规范化内容，而存在类别的调整。比如，对于待识别对象为身份证的情况，第一识别结果可以包括身份证正面记载的姓名、出生日期、居住地、身份证号、人像等内容中的一项或是多项。当然，第一识别结果还可以是基于上述内容得到的进一步的识别结果，比如，通过对识别到的身份证号进行进一步识别，以确定该身份证所属人物的出生地等数据。
通过上述公开实施例可以看出，第一识别结果可以是待识别对象中记录的内容，比如记录的文本、图片或是用以表明待识别对象的类别的标识等。在一个示例中，在待识别对象为证卡对象的情况下，第一识别结果可以是证卡对象中包含的文本，比如身份证上的地址，银行卡上的银行卡号码等。后续各公开实施例均以第一识别结果为文本为例进行阐述，第一识别结果为图片或是标识等其他类型的情况，可以根据后续各公开实施例进行灵活扩展，不再进行详细说明。
另外，第二设备如何根据待识别信息来得到第一识别结果，其具体的实现方式可以参见后续各公开实施例。
在本公开实施例中，通过获取待识别对象的图像序列，并从图像序列中选定至少一帧图像作为待识别图像，从而根据选定的待识别图像确定待识别信息并向第二设备发送待识别信息，以使第二设备根据待识别信息来得到第一识别结果。通过上述过程，可以在信息识别的过程中，首先对待识别对象自动选帧，再通过第二设备进行识别。由于自动选帧可以选取图像质量最优的图像，使得用于识别的图像比起直接获取的图像具有更高的图像精度和识别效果，因此可以使得最终得到的第一识别结果更为准确，同时也可以降低信息识别过程的失败率，从而提升信息识别过程中的用户体验。
第一设备通过步骤S11，利用任意一种方式获取了待识别对象的图像序列后，可以通过步骤S12从图像序列中确定至少一帧图像来作为待识别图像。具体如何确定待识别图像，这一过程可以根据实际情况灵活决定。在一种可能的实现方式中，步骤S12可以包括：
根据图像序列的识别状态和/或图像质量，选定至少一帧图像，作为待识别图像。
其中，识别状态可以指示识别出的图像序列中的图像内容的状态，例如，所识别的 图像内容的置信度等。图像内容的识别状态越优，说明该图像中包含的可以被识别的内容越全面与准确。即，根据识别状态确定待识别图像的过程，可以理解为，根据图像序列中不同帧图像之间识别状态的优劣来确定待识别图像。基于该待识别图像得到的第一识别结果，可以更加全面，具有更高的精度。
图像质量可以是图像序列中各帧图像的质量，图像质量的高低可以通过预设的质量指标来判断。例如，质量指标可以包括清晰度、完整度、强光情况、暗光情况以及遮挡情况等中的一者或多者，本文并不对此进行限制。图像质量越高，说明该帧图像越容易被识别，识别的结果也更加准确。即，根据图像质量确定待识别图像的过程，可以理解为，根据图像序列不同帧图像之间图像质量的高低来确定待识别图像。基于该待识别图像得到的第一识别结果，同样可以更为精确，且具有较优的识别效果。
根据识别状态或是图像质量来确定待识别图像的具体实现方式同样可以根据实际情况灵活选择，可以参考后续各公开实施例。
通过上述公开实施例可以看出，从图像序列中确定待识别图像，可以包含有基于图像序列的识别状态进行确定以及基于图像序列的图像质量进行确定两种方式。在一种可能的实现方式中，在确定待识别图像过程中，可以仅根据识别状态或图像质量来确定，即第一设备自动通过其中的一种方式来确定待识别图像。在一种可能的实现方式中，也可以同时根据识别状态和图像质量来确定，即第一设备可以结合两种方式来确定待识别图像。具体地，在一个示例中，第一设备可以分别通过基于识别状态和图像质量确定相应的待识别图像，然后将通过这两种方式确定的这些图像均作为最终的待识别图像进入到步骤S13。在一个示例中，第一设备也可以分别通过基于识别状态和图像质量确定相应的待识别图像后，再从中以一定的方式或是随机选择某一帧或某几帧作为最终的待识别图像进入到步骤S13。在一个示例中，第一设备也可以向用户提供根据识别状态确定待识别图像和根据图像质量确定待识别图像这两种选项，即让用户选择以哪种方式来确定待识别图像。
通过基于图像序列的识别状态和/或图像质量来确定待识别图像，一方面可以确保待识别图像的质量，使得基于此待识别图像最终得到的第一识别结果更加准确，具有更高精度，另一方面也可以在不同的情况下灵活选择不同的待识别图像确定方式，增加了信息识别方法的灵活性。
在一种可能的实现方式中，根据图像序列的识别状态，选定至少一帧图像，作为待识别图像可以包括：对图像序列中至少部分图像分别进行第二信息识别，得到一个或多个第二识别结果。根据一个或多个第二识别结果，得到参考识别结果。分别将每个第二识别结果与参考识别结果进行比较，得到每个第二识别结果的置信度。根据置信度，将全部第二识别结果的至少一部分对应的各图像，确定为待识别图像，其中，确定出的待识别图像的置信度高于图像序列中的非待识别图像的置信度。非待识别图像即为不用于信息识别的图像。
其中，第二信息识别可以是由第一设备对待识别对象的图像序列进行信息识别操作以确定待识别信息。上述公开实施例中提出过，第二设备可以根据待识别信息得到第一识别结果，说明第二设备可以根据待识别信息进行相应的信息识别操作。为了将不同实施主体的信息识别操作进行区分，在本公开实施例中，将第二设备根据待识别信息执行的操作记为第一信息识别，得到的结果记为第一识别结果，将第一设备对图像序列中至少部分图像执行的操作记为第二信息识别，得到的结果记为第二识别结果，第一信息识别与第二信息识别的具体识别方式可以相同，也可以不同。在一种可能的实现方式中， 由于第二信息识别主要为了确定待识别图像而非确定最终的识别结果，因此，第二信息识别可以与第一信息识别不同，第二信息识别可以选择识别精度较低的识别方式，第一信息识别可以选择识别精度较高的识别方式。在一个示例中，由于信息识别可以是对文本信息进行识别，因此第二信息识别和第一信息识别均可以通过光学字符识别(OCR，Optical Character Recognition)来实现。由于第二信息识别的识别精度可以无需达到第一信息识别的识别精度要求，因此，在本公开示例中，第二信息识别中使用的OCR模型规模可以小于第一信息识别中使用的OCR模型规模，从而既可以确保第二信息识别的实现，又可以保证第二信息识别的实现速度，继而提升整个信息识别过程的速度。
通过上述过程可以看出，第一设备在进行第二信息识别的过程中，可以对图像序列中的每帧图像均进行第二信息识别，也可以对图像序列中的部分图像进行第二信息识别，具体选择哪些帧图像来进行第二信息识别，可以根据实际情况灵活决定，在本公开实施例中不做限制。后续各公开实施例均以对图像序列中的每帧图像进行第二信息识别为例进行说明，其他情况可以灵活进行扩展。在一种可能的实现方式中，在对图像序列中每帧图像进行第二信息识别的情况下，可以分别得到每帧图像的第二识别结果。由于图像序列包括至少两帧图像，因此相应地可以得到至少两个第二识别结果，因此可以基于至少两个第二识别结果来得到参考识别结果。其中，参考识别结果可以是基于至少两个第二识别结果所整合的较为完整的识别结果。举例来说，在对多帧图像进行第二信息识别得到的多个第二识别结果中，可能有些第二识别结果缺少待识别内容的前半部分，有些第二识别结果缺少待识别内容的后半部分，有些第二识别结果缺少待识别内容的某个或某些字段。因此，当将这些第二识别结果进行统计后，可以恢复出一个较为完整和准确的识别结果，这一较为完整和准确的识别结果则可以作为参考识别结果。具体的恢复方式在本公开实施例中不做限制，可以根据实际情况灵活选择。在一个示例中，可以通过遍历每个第二识别结果，统计其中重复出现的识别内容，并根据重复出现的识别内容的位置进行整合，从而确定出参考识别结果。
在确定了参考识别结果后，可以将每个第二识别结果与参考识别结果进行比较，得到每个第二识别结果的置信度。其中，置信度代表了每个第二识别结果与参考识别结果的重合程度。在识别的是文本信息的情况下，置信度可以代表每个第二识别结果与参考识别结果的文本重合准确率，置信度越高，表明第二识别结果与参考识别结果越接近。
在得到了每个第二识别结果的置信度后，可以根据置信度的大小确定待识别图像。由于置信度越高表明第二识别结果与参考识别结果越接近，因此，置信度越高的第二识别结果所对应的图像，其包含的识别信息越全面，基于此图像进行信息识别的精度也越高。上述公开实施例中已经提出，待识别图像可以为一帧也可以为多帧，因此可以将置信度最高的一帧或多帧图像来作为待识别图像。或者，可以设置一个置信度阈值，将置信度高于该阈值的第二识别结果对应的图像，确定为待识别图像。或者，可以将置信度排名前20％的第二识别结果对应的图像，确定为待识别图像。这些确定待识别图像的示例并不具有限制性，还可以采用其它方式或其它数值。
通过上述过程，无论图像序列是在何种质量的摄像头、光线环境下所获得，均可以从中选择出相对较优的图像作为待识别图像，从而既可以确保信息识别过程的顺利实现，又可以尽可能的提升信息识别结果的准确性，并大大提升了信息识别方法的包容性。
在一种可能的实现方式中，根据图像序列的图像质量，选定至少一帧图像，作为待识别图像，可以包括：分别获取图像序列中至少部分图像在至少一个衡量维度下的图像质量。将图像序列中图像质量大于对应衡量维度下的阈值的至少一帧图像，确定作为待 识别图像。
其中，获取图像序列中图像的图像质量的实现过程中，可以分别获取图像序列中每帧图像的图像质量，也可以仅获取图像序列中部分帧图像的图像质量，具体选择哪些帧以及如何选择，可以根据实际情况灵活决定，在本公开实施例中不做限制。后续各公开实施例均以获取图像序列中每帧图像的图像质量为例进行说明，其他情况的实现方式可以参考下述各公开实施例进行灵活扩展。
上述公开实施例中已经提出，图像质量可以是图像序列中图像的质量，对于任意一帧图像来说，评判该图像的质量如何，可以具有不同的评判标准，比如可以从清晰度或是完整度等不同的角度来分别评判。因此，随着评判标准的不同，可以在不同衡量维度下分析图像质量。具体的，通过某一衡量维度或是某几种衡量维度，以及每个衡量维度具体对应何种评判标准来分析图像质量，均可以根据实际情况进行灵活设定。后续在根据图像质量与阈值的比较确定待识别图像时，可以不用考虑全部的衡量维度，而是在选定衡量维度下比较图像质量。在一种可能的实现方式中，衡量维度可以包括：清晰度、完整度、强光情况、暗光情况以及遮挡情况中的一个或两个以上。其中，清晰度可以指示图像是否存在对焦模糊、运动模糊等导致文字或图像识别不清的情况；完整度可以指示图像中待识别对象(如证件)的边角是否全部处于图像范围之内等；强光情况可以指示图像是否存在过曝或强烈反光等情况；暗光情况可以指示图像中待识别对象(如证件)是否存在整体或局部亮度过暗，导致文字或图像无法识别等情况；遮挡情况可以指示图像中待识别对象(如证件)是否有被其他的物体遮挡等情况。在实际应用中，可以只考虑其中某一衡量维度，也可以同时考虑其中多个衡量维度。在一个示例中，可以同时考虑清晰度、完整度、强光情况、暗光情况和遮挡情况这五个衡量维度。在这种情况下，可以分别获取图像序列中每帧图像在这五个衡量维度下的图像质量，然后对每帧图像，分别考虑其清晰度质量是否大于对应的清晰度阈值，完整度质量是否大于对应的完整度阈值，强光质量是否大于对应的强光阈值，暗光质量是否大于对应的暗光阈值以及遮挡质量是否大于对应的遮挡阈值。当存在某一帧或某几帧图像在这五个类别下均满足大于对应阈值的要求时，可以将满足要求的图像作为选定图像。在一个示例中，在衡量维度包含上述五个维度的情况下，也可以获取图像在其中部分维度的图像质量，举例来说，可以仅获取图像在清晰度、完整度以及遮挡情况这三个衡量维度下的质量。在这种情况下，后续确定待识别图像，可以只考虑每帧图像的清晰度质量、完整度质量以及遮挡情况质量是否分别大于对应的阈值，而省略强光质量与暗光质量的比较情况。
进一步地，在一种可能的实现方式中，当满足要求的图像数量大于需要的待识别图像数量的情况下，可以进一步比较这些满足要求的图像的图像质量，选定综合图像质量较高的N个图像作为待识别图像，其中N为需要的待识别图像数量。具体地，综合质量的计算方式可以根据实际情况进行设定，根据实际情况选择具体计算方式即可。比如可以为每个衡量维度下的图像质量设置一个权重，从而计算出每帧图像的加权平均图像质量，将加权平均图像质量作为综合图像质量等。在一种可能的实现方式中，也可能没有满足要求的图像，为了确保信息识别过程的顺利进行，也可以通过比较这些图像的图像质量，选定综合图像质量较高的N个图像作为待识别图像。
在一种可能的实现方式中，也可以将获取待识别对象的图像序列的过程与根据图像质量确定选定图像的过程同时进行，即可以一边采集待识别对象的图像序列，一边对已经采集到的图像进行图像质量的评判。当已经采集到的图像中包括有满足各个衡量维度下的图像质量要求的图像的情况下，可以将这样的图像作为待识别图像，并停止继续采 集待识别对象的图像序列。这种情况下，可能获取到的图像序列中的最后一帧图像即为待识别图像。
另外，每个衡量维度下的图像质量对应的阈值，在本公开实施例中也不做限制，可以根据实际情况进行设定，不同衡量维度下图像质量的阈值可以相同，也可以不同。
通过分别获取图像序列中至少部分图像在至少一个衡量维度下的图像质量，从而将在每个衡量维度下的图像质量均大于对应阈值的至少一帧图像作为待识别图像。通过上述过程，可以从图像序列中选出高质量的待识别图像，从而使得基于此待识别图像进行信息识别得到的第一识别结果具有更高的准确度，大大提升信息识别的准确性。
在一种可能的实现方式中，在通过步骤S12确定了待识别图像以后，还可以对待识别图像进行存档。因此，在一种可能的实现方式中，在确定待识别图像之后，本公开实施例提出的方法还可以包括：保存待识别图像中的至少一帧。
其中，待识别图像的保存位置不受限定，既可以保存在第一设备中，也可以保存在第二设备中，也可以同时保存在第一设备和第二设备之中。这样，在后续还有需要使用待识别对象的图像或是需要对待识别对象再次进行识别的情况下，可以直接读取保存的待识别图像，大大提升了效率和用户的体验程度。待识别图像的保存帧数也不受限制，可以根据第一设备以及第二设备的存储空间大小灵活决定。在一种可能的实现方式中，在待识别图像为多帧的情况下，可以保存全部的待识别图像，也可以从中选择一帧或几帧进行保存；在一种可能的实现方式中，在待识别图像为一帧的情况下，可以直接保存该待识别图像。
进一步地，由于根据识别状态确定待识别图像和根据图像质量确定待识别图像可以同时实现，也可以单独实现，相应的，保存待识别图像中的至少一帧可以仅在根据识别状态确定待识别图像后实现，即仅保存根据识别状态确定的待识别图像，也可以仅在根据图像质量确定待识别图像后实现，即仅保存根据图像质量确定的待识别图像，也可以同时在根据识别状态和图像质量确定待识别图像后实现，即可以保存根据上述两种确定方式综合确定的待识别图像。在一种可能的实现方式中，由于基于图像质量确定的待识别图像，相对于基于识别状态确定的待识别图像来说，具有较优的图像质量，因此，可以仅保存通过图像质量所确定的待识别图像。
通过上述各公开实施例还可以看出，在确定了待识别图像后，还可以通过步骤S13，来根据待识别图像确定待识别信息，在一种可能的实现方式中，步骤S13可以包括：
对待识别图像进行第一加密处理和/或签名处理，得到待识别信息。
其中，第一加密处理与签名处理的实现方式均不受限定，可以根据实际情况灵活选择。在一种可能的实现方式中，第一加密处理可以包括：对待识别图像进行编码加密，得到第一加密信息；签名处理可以包括：将第一设备的签名信息添加至选定图像。其中，对待识别图像以何种编码方式进行加密，在本公开实施例中不做限制，任何加密的方法均可以作为第一加密处理的实现方式。第一设备的签名信息可以是包含有第一设备身份标识的信息，其具体的信息内容和形式在本公开实施例中也不做限制。在一个示例中，第一设备的签名信息可以是第一设备中软件开发工具包(SDK，Software Development Kit)中的签名信息。同样地，将第一设备的签名信息添加到待识别图像的具体添加位置和方式，在本公开实施例中也不做限定，根据实际情况灵活选择即可。
可以对待识别图像进行第一加密处理与签名处理的一种或两种。当对待识别图像进行第一加密处理与签名处理这二者时，这两种处理可以同时进行，也可以按照一定的顺序先后进行，根据实际情况灵活选择即可。在一种可能的实现方式中，可以依次对待识 别图像进行第一加密处理和签名处理。即，先将待识别图像按照一定的加密形式进行重新编码，然后将编码加密得到的信息与第一设备的签名信息进行打包，从而得到待识别信息。在一种可能的实现方式中，也可以依次对待识别图像进行签名处理和第一加密处理。即，先将待识别图像与第一设备的签名信息进行打包，然后将打包得到的信息按照一定的加密形式进行重新编码，从而得到待识别信息。
通过第一加密处理，在待识别对象为身份证件或银行卡等具有较高安全需求的对象的情况下，可以有效减小证卡信息外泄的可能性，从而可以提升信息识别过程的安全性；而通过签名处理，则可以在待识别信息中包含有第一设备的设备信息，从而便于第二设备在进行信息识别时对待识别信息进行权限验证，减小待识别信息被篡改的可能性，进一步提升了信息识别过程的安全性，也提升了信息识别结果的准确性。
在得到了待识别信息后，可以通过步骤S14，向第二设备发送待识别信息，具体的发送方式根据第一设备和第二设备之间的连接关系灵活确定。在一种可能的实现方式中，当第一设备为终端设备，而第二设备为服务器的情况下，第一设备可以通过第一设备与第二设备之间连接的网络，向第二设备上传待识别信息。
在一种可能的实现方式中，在向第二设备发送了待识别信息后，还可以接收第二设备所反馈的识别信息。由于第二设备可以根据待识别信息得到第一识别结果，因此，识别信息可以是与第一识别结果相关的信息。识别信息的具体实现方式可以参见下述各公开实施例。第一设备接收识别信息的方式不受限定，同样可以根据第一设备与第二设备之间的通信方式灵活决定，在此不再赘述。
第一设备在接收了识别信息后，可以将识别信息显示给用户，也可以在显示的同时保存该识别信息，具体如何应用该识别信息，可以根据第一设备的实际需求灵活决定，在本公开实施例中不做限定。
图2示出根据本公开一实施例的信息识别方法的流程图，该方法可以应用于第二设备，第二设备的实现方式可以参考上述各公开实施例，在此不再赘述。
如图2所示，在一种可能的实现方式中，所述信息识别方法可以应用于第二设备，包括：步骤S21，接收待识别对象的待识别信息。步骤S22，对待识别信息进行第一信息识别，得到第一识别结果。步骤S23，根据第一识别结果，向第一设备发送识别信息。
其中，待识别信息与上述公开实施例中提到的待识别信息一致，在此不再赘述。第二设备接收待识别信息的方式可以根据第一设备发送待识别信息的方式灵活决定，在本公开实施例中不做限制。第二设备在接收了待识别信息后，可以通过步骤S22来进行第一信息识别，具体的识别过程可以参见下述各公开实施例。在第二设备进行第一信息识别得到第一识别结果后，可以通过步骤S23，根据第一识别结果来向第一设备发送识别信息，其中，识别信息可以根据第一识别结果得到。在一种可能的实现方式中，可以直接将第一识别结果作为识别信息。在一种可能的实现方式中，也可以根据第一识别结果，额外进行一些其他的处理或内容，来得到识别信息，具体的识别信息的内容和生成方式可以参见后续各公开实施例。
通过接收待识别对象的待识别信息，对待识别信息进行第一信息识别来得到第一识别结果，从而基于第一识别结果向第一设备发送识别信息。通过第一设备与第二设备之间的交互配合，有效地实现了对待识别对象的信息识别。由于第一设备实现选帧功能，第二设备实现识别功能，两个设备均实现单一功能，对计算要求较低，因此，在应用中通过两个设备的配合，可以具有更快的处理速度和更强的计算效果，从而大大提升了整个信息识别过程的效率和精度。
第二设备进行第一信息识别的具体方式可以根据实际情况灵活决定，在一种可能的实现方式中，步骤S22可以包括：步骤S221，获取待识别信息包括的待识别图像。步骤S222，对待识别图像进行防伪检测，得到检测结果。步骤S223，在检测结果为通过的情况下，对待识别图像进行第一信息识别，得到第一识别结果。
上述公开实施例中已经提出，可以直接将待识别图像作为待识别信息，也可以对待识别图像进行一定的处理来得到待识别信息。因此，随着待识别信息的生成方式的不同，步骤S221的实现方式也可以随之发生变化，详见下述各公开实施例。在得到了待识别图像后，可以通过步骤S222对待识别图像进行防伪检测。在待识别对象为证卡对象等具有较高安全需求的对象的情况下，可能会有某些用户将待识别对象的复印件或是翻拍件等作为待识别对象来进行信息识别。为了提升信息识别的安全，需要对待识别图像进行防伪检测，以减少这类情况的发生，具体的防伪检测方式可以根据实际情况进行选择，参考后续各公开实施例。在防伪检测通过后，才可以对待识别图像进行第一信息识别，如果防伪检测失败，则可以向第一设备反馈检测失败或是报警提示等信息，以进一步提升信息识别过程的安全性。
通过获取待识别信息包括的待识别图像，然后对待识别图像进行防伪检测，得到检测结果，在检测结果为通过的情况下对待识别图像进行第一信息识别，得到第一识别结果。通过上述过程，可以有效提升整个信息识别过程的安全性。
步骤S221的实现方式可以根据实际情况灵活选择。在一种可能的实现方式中，在待识别信息为待识别图像的情况下，可以直接从待识别信息中读取待识别图像。在一种可能的实现方式中，当待识别信息为对待识别图像进行第一加密处理所得到的信息的情况下，可以通过对待识别信息进行解密来得到待识别图像。在一种可能的实现方式中，在待识别信息为通过对待识别图像进行第一加密处理和签名处理所得到的信息的情况下，步骤S221可以包括：步骤S2211，获取待识别信息包括的签名信息。步骤S2212，在获取的签名信息与第一设备的签名信息匹配的情况下，对待识别信息包括的第一加密信息进行解密，得到待识别图像。
在一种可能的实现方式中，第二设备可以首先根据待识别信息中包括的签名信息，对待识别信息进行权限验证。当该签名信息与第一设备的签名信息匹配的情况下，可以说明待识别信息在发送到第二设备的过程中，没有被其他的设备或用户进行篡改，即待识别信息中的待识别图像是可以被用于信息识别的图像。继而，第二设备可以对第一加密信息进行解密，从而还原出第一设备确定的待识别图像。解密的方式可以根据加密的方式灵活确定，在本公开实施例中不做限定。
进一步地，若待识别信息中包括的签名信息与第一设备的签名信息不匹配，则说明待识别信息可能被篡改过，此时第二设备可以停止对待识别图像的获取，而是向第一设备反馈匹配失败或是报警提示等信息，以进一步提升信息识别过程的安全性。
通过上述过程，可以实现对待识别信息的权限验证，在待识别信息被篡改的情况下一方面可以发出报警提示，另一方面又可以减少无意义的解密与识别过程，同时提升信息识别的安全性和效率。
同样地，步骤S222的实现方式也不受限定，即对待识别图像进行防伪检测的方式不受限定，可以根据实际情况灵活选择。在一种可能的实现方式中，可以基于待识别对象的复印件或是翻拍件等具有的独特特征，来确定待识别图像是否为复印件或翻拍件对应的帧图像，比如翻拍件可能存在反光等情况，复印件可能在色彩上与原件具有显著差异等。在一种可能的实现方式中，步骤S222可以包括：步骤S2221，对待识别图像 进行分类，得到待识别图像的分类结果。步骤S2222，在分类结果指示待识别图像为通过对待识别对象进行拍摄得到的图像的情况下，将检测结果记录为通过。和/或，步骤S2223，在分类结果指示待识别图像为通过对待识别对象的复印件或翻拍件进行拍摄得到的图像的情况下，将检测结果记录为失败。
在一种可能的实现方式中，可以通过对待识别图像进行分类，来确定待识别图像的类型，从而实现对待识别图像的防伪检测。如何对待识别图像进行分类，其实现方式不受限定。在一种可能的实现方式中，可以通过分类神经网络模型来对待识别图像进行分类，其中，分类神经网络模型的具体实现方式不受限制。在一种可能的实现方式中，可以建立一个初始的神经网络模型，然后将大量的经过拍摄所得到的待识别对象的图像、待识别对象的复印件图像和待识别对象的翻拍件图像作为训练样本，对该初始的神经网络模型进行训练，从而得到一个训练好的分类神经网络模型。将待识别图像输入到该分类神经网络模型后，可以输出一个概率值，用以表明该待识别图像为待识别对象的原件图像的概率。当这一概率值大于设定的概率阈值的情况下，可以表明该待识别图像为待识别对象的原件图像，否则表明该待识别图像为待识别对象的复印件图像或翻拍件图像，具体的概率阈值可以根据实际情况进行灵活设定，在本公开实施例中不做限制。
对待识别图像进行分类得到分类结果，在分类结果指示待识别图像是对待识别对象进行拍摄所得到的图像的情况下，将检测结果记录为通过，否则记录为失败。通过这一过程，可以利用分类方式来实现对待识别图像的防伪检测，既具有较高的检测效率，又具有较高的检测精度，从而大大提升了整个信息识别过程的准确性和速度。
在检测结果为通过的情况下，可以进一步对待识别图像进行第一信息识别。上述公开实施例中已经提到，第一信息识别的方式可以与第二信息识别的方式相同，也可以不同，根据实际情况灵活选择即可。在一种可能的实现方式中，步骤S223可以包括：通过光学字符识别OCR模型，对待识别图像进行OCR识别，得到第一识别结果。在一种可能的实现方式中，由于第一信息识别的目的在于识别出待识别图像中包含的相应信息，具有较高的识别精度的需求，因此，在本公开实施例中，可以通过较大规模的OCR模型来实现第一信息识别，来提升信息识别的精度和准确性。
由于各种因素的存在，所得到的第一识别结果可能与待识别对象本身包含的内容一致，也可能存在一定的偏差。在存在偏差的情况下，还可以对第一识别结果进行进一步地校验。因此，在一种可能的实现方式中，步骤S22还可以包括：步骤S224，判断第一识别结果是否与预设规则匹配，得到判断结果。
其中，预设规则可以是根据待识别对象中的信息的特点所确定的某些验证规则。在一个示例中，当待识别对象为身份证，第一识别结果为身份证号码的情况下，由于身份证号码的后四位遵循一定的编码规则，因此可以将这一编码规则作为预设规则，来判断第一识别结果的身份证号码是否为真实存在的号码；同样地，当待识别对象为银行卡，第一识别结果为银行卡号码的情况下，也可以将银行卡号码的编码规则作为预设规则。当待识别对象为其他类型的情况下，预设规则可以参考上述各公开实施例进行类比扩展，在此不再一一列举。
当第一识别结果符合预设规则的情况下，可以将判断结果记录为匹配通过，当第一识别结果不符合预设规则的情况下，则可以将判断结果记录为匹配失败。通过上述过程，可以进一步对第一识别结果进行验证，从而在第一识别结果与预设规则不匹配的情况下，发出一定的提示或预警，以便于用户确认该识别结果是否准确，是否需要重新识别等。
在得到了第一识别结果后，可以基于该第一识别结果，通过步骤S23向第一设备发送识别信息，步骤S23的实现方式可以根据实际情况灵活决定，在一种可能的实现方式中，步骤S23可以包括：步骤S231，对第一识别结果进行第二加密处理，或是对第一识别结果和判断结果进行第二加密处理，得到第二加密信息；步骤S232，将第二加密信息作为识别信息，并向第一设备发送。
在一种可能的实现方式中，可以直接对第一识别结果进行第二加密处理，来得到第二加密信息，从而将第二加密信息作为识别信息发送至第一设备。其中，第二加密处理与第一加密处理中的“第一”和“第二”仅用于区别执行加密处理的主体和对象不同，即第一加密处理是第一设备对待识别图像进行的加密，而第二加密处理是第二设备对第一识别结果的加密，而不限制加密方式是否相同，即第一加密处理和第二加密处理的加密规则可以相同也可以不同，根据实际情况灵活选择即可。
在一种可能的实现方式中，由于将第一识别结果与预设规则进行了匹配来得到判断结果，因此，可以将判断结果与第一识别结果进行打包，并将打包得到的信息一并进行第二加密处理来得到第二加密信息，将该第二加密信息作为识别信息发送至第一设备，从而可以便于用户或第一设备根据判断结果来做出是否需要再次进行信息识别的决定等。
在一种可能的实现方式中，本公开实施例提出的应用于第二设备的信息识别方法还可以包括：保存待识别图像中的至少一帧。
上述公开实施例中已经提出过，在第一设备进行信息识别的过程中，就可以保存待识别图像中的至少一帧，同理，由于第二设备可以从接收的待识别信息中恢复出待识别图像，因此，第二设备也可以对待识别图像中的至少一帧进行保存。
与上述公开实施例中相同，第二设备在保存待识别图像时，具体保存的是一帧待识别图像还是多帧待识别图像，可以根据实际情况灵活决定，在此不再赘述。由于待识别图像可能是第一设备基于识别状态所选定的图像，也可能是第一设备基于多个衡量维度下图像质量所选定的图像，随着选定方式的不同，待识别图像的质量也可以产生相应的变化，因此，在一种可能的实现方式中，第二设备可以根据待识别图像的选定方式的不同，而选择不同的保存方式，具体如何选择，可以根据实际情况进行灵活设定，不局限于下述公开实施例。
在一种可能的实现方式中，由于基于识别状态所确定的待识别图像可能在完整度或清晰度上有所欠缺，因此第二设备可以选择仅保存通过基于多个衡量维度下图像质量所确定的待识别图像，来提升第二设备中保存的待识别图像的质量。在一种可能的实现方式中，当第二设备获得的待识别图像的数量为多个的情况下，也可以对这多个待识别图像再进行一次图像质量筛选，从而将图像质量最高的图像保存在第二设备中，第二设备中进行图像质量筛选的方式可以与第一设备中的筛选方式相同，也可以不同，根据实际情况进行灵活选择即可。
通过在第二设备中保存待识别图像中的至少一帧，可以实现待识别图像的远程保存，在需要再次对待识别对象进行识别或需要使用待识别对象的图像的其他情况下，可以直接从第二设备中调取待识别图像，减少用户的操作，提升用户体验。
图3示出根据本公开实施例的信息识别装置30的框图。如图3所示，所述装置30可以包括：图像序列获取模块31，用于获取待识别对象的图像序列，图像序列包括至少两帧图像。确定模块32，用于从图像序列中确定至少一帧图像，作为待识别图像。待识别信息生成模块33，用于根据待识别图像，确定待识别信息。待识别信息发送模块 34，用于向第二设备发送待识别信息，以使第二设备根据待识别信息得到第一识别结果。
在一种可能的实现方式中，确定模块用于：根据图像序列的识别状态和/或图像质量，从图像序列中确定至少一帧图像，作为待识别图像。
在一种可能的实现方式中，确定模块进一步用于：对图像序列中至少部分图像分别进行第二信息识别，得到一个或多个第二识别结果；根据一个或多个第二识别结果，得到参考识别结果；分别将每个第二识别结果与参考识别结果进行比较，得到每个第二识别结果的置信度；根据置信度，将全部第二识别结果的至少一部分对应的各图像，确定为待识别图像，其中，确定出的待识别图像的置信度高于图像序列中非待识别图像的置信度。
在一种可能的实现方式中，确定模块进一步用于：分别获取图像序列中至少部分图像在至少一个衡量维度下的图像质量；将图像序列中图像质量大于对应衡量维度下的阈值的至少一帧图像，确定为待识别图像。
在一种可能的实现方式中，衡量维度包括清晰度、完整度、强光情况、暗光情况以及遮挡情况中的至少一项。
在一种可能的实现方式中，所述装置30还包括第一保存模块，第一保存模块用于：保存待识别图像中的至少一帧。
在一种可能的实现方式中，待识别信息生成模块用于：对待识别图像进行第一加密处理和/或签名处理，得到待识别信息。
在一种可能的实现方式中，第一加密处理包括：对待识别图像进行编码加密，得到第一加密信息；签名处理包括：将第一设备的签名信息添加至待识别图像。
在一种可能的实现方式中，待识别对象包括证卡对象和/或表单对象；第一识别结果包括待识别对象中记录的文本、字符、标识、图标、以及图片中的至少一项。
图4示出根据本公开实施例的信息识别装置的框图。如图所示，所述装置40可以包括：接收模块41，用于接收待识别对象的待识别信息。识别模块42，用于对待识别信息进行第一信息识别，得到第一识别结果。识别信息发送模块43，用于根据第一识别结果，向第一设备发送识别信息。
在一种可能的实现方式中，识别模块用于：获取待识别信息包括的待识别图像；对待识别图像进行防伪检测，得到检测结果；在检测结果为通过的情况下，对待识别图像进行第一信息识别，得到第一识别结果。
在一种可能的实现方式中，识别模块进一步用于：获取待识别信息包括的签名信息；在获取的签名信息与第一设备的签名信息匹配的情况下，对待识别信息包括的第一加密信息进行解密，得到待识别图像。
在一种可能的实现方式中，识别模块进一步用于：对待识别图像进行分类，得到待识别图像的分类结果；在分类结果指示待识别图像为通过对待识别对象进行拍摄得到的图像的情况下，将检测结果记录为通过；和/或，在分类结果指示待识别图像为通过对待识别对象的复印件或翻拍件进行拍摄得到的图像的情况下，将检测结果记录为失败。
在一种可能的实现方式中，识别模块进一步用于：通过光学字符识别OCR模型，对待识别图像进行OCR识别，得到第一识别结果。
在一种可能的实现方式中，识别模块还用于：判断第一识别结果是否与预设规则匹配，得到判断结果。
在一种可能的实现方式中，识别信息发送模块用于：对第一识别结果进行第二加密处理，或是对第一识别结果和判断结果进行所述第二加密处理，得到第二加密信息； 将第二加密信息作为所述识别信息，并向第一设备发送。
在一种可能的实现方式中，装置40还包括第二保存模块，第二保存模块用于：保存待识别图像中的至少一帧。
在一种可能的实现方式中，本公开实施例还公开了一应用示例，该应用示例提出了一种信息识别系统，基于此信息识别系统，可以在线上贷款、租房或是会员注册等业务流程中，实现对身份证信息的识别与采集。
图5示出根据本公开一应用示例的示意图，如图5所示，信息识别系统主要由前端的第一设备和后端的第二设备共同构成，其中，前端的第一设备(以下简称前端)可以是手机设备、平板设备、具有摄像头的笔记本或电脑设备等，而后端的第二设备(以下简称后端)则可以是云端服务器。从图中可以看出，该信息识别系统实现身份证信息的识别与采集的完整过程可以概况为：第一设备可以通过获取待识别对象的图像序列，从图像序列中选定至少一帧作为待识别图像，然后根据待识别图像确定待识别信息从而将待识别信息发送至第二设备，第二设备在接收到待识别信息后，可以对待识别信息进行第一信息识别来得到第一识别结果，并根据第一识别结果向第一设备发送识别信息。
具体地，该信息识别系统对身份证信息进行识别采集的过程可以为：
前端可以打开摄像头对待识别的证件进行持续扫描选帧，直到选出满足条件的一帧作为待识别图像后，将选出的该待识别图像传输到后端进行后端处理。传输过程中可以将待识别图像进行编码加密，同时附带了前端SDK中的签名信息，用于权限验证，以确保数据传输过程中待识别图像不被篡改。
后端可以在接收到前端发送的待识别图像后，进行解密和权限校验，经由解密与权限校验确认该数据未被篡改后，进入识别流程。具体的识别流程可以为，首先对证件图像通过神经网络模型进行防伪检测，从而判断证件是否为原件；其次对身份证进行文字识别，采用OCR技术识别出证件的各项字段；最后根据预设的逻辑，对文字识别的结果进行校验判断。识别完成后，将文字识别的结果及逻辑判断结果加密传输回前端，提供给用户。
进一步地，在本公开应用示例中，前端进行选帧的过程可以从以下两种方式中任选其一来实现。
方式一：基于前端对采集的图像序列的各帧的文本识别结果选帧
此种选帧方式，可以使用前端的OCR小模型，对摄像头采集到的每一帧图像进行文字识别，依据文字识别结果的置信度来判断帧的质量。具体地，前端对每帧图像进行文本识别，得到多个文本识别结果，之后将多个文本识别结果整合出一个标准识别结果，来对多帧图像进行筛选，即选出文本识别结果最接近标准识别结果的一帧图像作为待识别图像，交给后端处理。由于OCR小模型的准确率及字库规模有限，仍然需要将待识别图像传输到后端进行OCR大模型的识别。通过此种选帧方式，可以具有较好的包容性，在不同质量的摄像头、光线环境下都可以有较好的选帧结果。
方式二：基于图像质量选帧
此种选帧方式，可以从清晰度、完整度、强光情况、暗光情况、遮挡情况这五个维度，分别进行证件图像的质量检测。在一个示例中，这五种维度的定义分别为：
清晰度，主要描述图像是否存在对焦模糊、运动模糊等导致文字或图像识别不清的情况；完整度，主要描述证件的边角是否全部处于图像范围之内；强光情况，主要描述证件是否存在过曝或强烈反光的情况；暗光情况，主要描述证件是否存在整体或局部亮度过暗，导致文字或图像无法识别的情况；遮挡情况，主要描述图像中证件是否有被其 他的物体遮挡。
在前端扫描的过程中，可以对每一帧图像进行上述五个维度的检测，并将检测结果与预设的阈值(该预设的阈值可根据实际情况进行调整)进行比较，在五个维度的检测结果均满足阈值条件的情况下，可以输出选帧结果，进行后续的后端识别。通过此种选帧方式，可以甄选出高质量的图像用于留底存档。
需要注意的是，上述应用示例提出的方法，除了可以应用于以上提到的场景以外，也可以应用于其他具有信息识别需求的线上业务等，如在线办卡或是特殊情况下的身份核验场景等，不局限于上述应用示例。
可以理解，本公开提及的上述各个方法实施例，在不违背原理逻辑的情况下，均可以彼此相互结合形成结合后的实施例，限于篇幅，本公开不再赘述。
本领域技术人员可以理解，在具体实施方式的上述方法中，各步骤的撰写顺序并不意味着严格的执行顺序而对实施过程构成任何限定，各步骤的具体执行顺序应当以其功能和可能的内在逻辑确定。
本公开实施例还提出一种计算机可读存储介质，其上存储有计算机程序指令，所述计算机程序指令被处理器执行时实现上述方法。计算机可读存储介质可以是易失性计算机可读存储介质或非易失性计算机可读存储介质。
本公开实施例还提出一种电子设备，包括：处理器；用于存储处理器可执行指令的存储器；其中，所述处理器被配置为上述方法。
在实际应用中，上述存储器可以是易失性存储器(volatile memory)，例如RAM；或者非易失性存储器(non-volatile memory)，例如ROM，快闪存储器(flash memory)，硬盘(Hard Disk Drive，HDD)或固态硬盘(Solid-State Drive，SSD)；或者上述种类的存储器的组合，并向处理器提供指令和数据。
上述处理器可以为ASIC、DSP、DSPD、PLD、FPGA、CPU、控制器、微控制器、微处理器中的至少一种。可以理解地，对于不同的设备，用于实现上述处理器功能的电子器件还可以为其它，本公开实施例不作具体限定。
电子设备可以被提供为终端、服务器或其它形态的设备。
基于前述实施例相同的技术构思，本公开实施例还提供了一种计算机程序，该计算机程序被处理器执行时实现上述方法。
图6是根据本公开实施例的一种电子设备800的框图。例如，电子设备800可以是移动电话，计算机，数字广播终端，消息收发设备，游戏控制台，平板设备，医疗设备，健身设备，个人数字助理等终端。
参照图6，电子设备800可以包括以下一个或多个组件：处理组件802，存储器804，电源组件806，多媒体组件808，音频组件810，输入/输出(I/O)接口812，传感器组件814，以及通信组件816。
处理组件802通常控制电子设备800的整体操作，诸如与显示，电话呼叫，数据通信，相机操作和记录操作相关联的操作。处理组件802可以包括一个或多个处理器820来执行指令，以完成上述的方法的全部或部分步骤。此外，处理组件802可以包括一个或多个模块，便于处理组件802和其他组件之间的交互。例如，处理组件802可以包括多媒体模块，以方便多媒体组件808和处理组件802之间的交互。
存储器804被配置为存储各种类型的数据以支持在电子设备800的操作。这些数据的示例包括用于在电子设备800上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器804可以由任何类型的易失性或非易失性存 储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
电源组件806为电子设备800的各种组件提供电力。电源组件806可以包括电源管理系统，一个或多个电源，及其他与为电子设备800生成、管理和分配电力相关联的组件。
多媒体组件808包括在所述电子设备800和用户之间的提供一个输出接口的屏幕。在一些实施例中，屏幕可以包括液晶显示器(LCD)和触摸面板(TP)。如果屏幕包括触摸面板，屏幕可以被实现为触摸屏，以接收来自用户的输入信号。触摸面板包括一个或多个触摸传感器以感测触摸、滑动和触摸面板上的手势。所述触摸传感器可以不仅感测触摸或滑动动作的边界，而且还检测与所述触摸或滑动操作相关的持续时间和压力。在一些实施例中，多媒体组件808包括一个前置摄像头和/或后置摄像头。当电子设备800处于操作模式，如拍摄模式或视频模式时，前置摄像头和/或后置摄像头可以接收外部的多媒体数据。每个前置摄像头和后置摄像头可以是一个固定的光学透镜系统或具有焦距和光学变焦能力。
音频组件810被配置为输出和/或输入音频信号。例如，音频组件810包括一个麦克风(MIC)，当电子设备800处于操作模式，如呼叫模式、记录模式和语音识别模式时，麦克风被配置为接收外部音频信号。所接收的音频信号可以被进一步存储在存储器804或经由通信组件816发送。在一些实施例中，音频组件810还包括一个扬声器，用于输出音频信号。
I/O接口812为处理组件802和外围接口模块之间提供接口，上述外围接口模块可以是键盘，点击轮，按钮等。这些按钮可包括但不限于：主页按钮、音量按钮、启动按钮和锁定按钮。
传感器组件814包括一个或多个传感器，用于为电子设备800提供各个方面的状态评估。例如，传感器组件814可以检测到电子设备800的打开/关闭状态，组件的相对定位，例如所述组件为电子设备800的显示器和小键盘，传感器组件814还可以检测电子设备800或电子设备800一个组件的位置改变，用户与电子设备800接触的存在或不存在，电子设备800方位或加速/减速和电子设备800的温度变化。传感器组件814可以包括接近传感器，被配置用来在没有任何的物理接触时检测附近物体的存在。传感器组件814还可以包括光传感器，如CMOS或CCD图像传感器，用于在成像应用中使用。在一些实施例中，该传感器组件814还可以包括加速度传感器，陀螺仪传感器，磁传感器，压力传感器或温度传感器。
通信组件816被配置为便于电子设备800和其他设备之间有线或无线方式的通信。电子设备800可以接入基于通信标准的无线网络，如WiFi，2G或3G，或它们的组合。在一个示例性实施例中，通信组件816经由广播信道接收来自外部广播管理系统的广播信号或广播相关人员信息。在一个示例性实施例中，所述通信组件816还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
在示例性实施例中，电子设备800可以被一个或多个应用专用集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理设备(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于 执行上述方法。
在示例性实施例中，还提供了一种非易失性计算机可读存储介质，例如包括计算机程序指令的存储器804，上述计算机程序指令可由电子设备800的处理器820执行以完成上述方法。
图7是根据本公开实施例的一种电子设备1900的框图。例如，电子设备1900可以被提供为一服务器。参照图7，电子设备1900包括处理组件1922，其进一步包括一个或多个处理器，以及由存储器1932所代表的存储器资源，用于存储可由处理组件1922执行的指令，例如应用程序。存储器1932中存储的应用程序可以包括一个或一个以上的每一个对应于一组指令的模块。此外，处理组件1922被配置为执行指令，以执行上述方法。
电子设备1900还可以包括一个电源组件1926被配置为执行电子设备1900的电源管理，一个有线或无线网络接口1950被配置为将电子设备1900连接到网络，和一个输入输出(I/O)接口1958。电子设备1900可以基于存储在存储器1932的操作系统进行操作，例如Windows ServerTM，Mac OS XTM，UnixTM,LinuxTM，FreeBSDTM或类似。
在示例性实施例中，还提供了一种非易失性计算机可读存储介质，例如包括计算机程序指令的存储器1932，上述计算机程序指令可由电子设备1900的处理组件1922执行以完成上述方法。
本公开提供了一种系统、方法和/或计算机程序产品。计算机程序产品可以包括计算机可读存储介质，其上载有用于使处理器实现本公开的各个方面的计算机可读程序指令。
计算机可读存储介质可以是可以保持和存储由指令执行设备使用的指令的有形设备。计算机可读存储介质例如可以是――但不限于――电存储设备、磁存储设备、光存储设备、电磁存储设备、半导体存储设备或者上述的任意合适的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括：便携式计算机盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、静态随机存取存储器(SRAM)、便携式压缩盘只读存储器(CD-ROM)、数字多功能盘(DVD)、记忆棒、软盘、机械编码设备、例如其上存储有指令的打孔卡或凹槽内凸起结构、以及上述的任意合适的组合。这里所使用的计算机可读存储介质不被解释为瞬时信号本身，诸如无线电波或者其他自由传播的电磁波、通过波导或其他传输媒介传播的电磁波(例如，通过光纤电缆的光脉冲)、或者通过电线传输的电信号。
这里所描述的计算机可读程序指令可以从计算机可读存储介质下载到各个计算/处理设备，或者通过网络、例如因特网、局域网、广域网和/或无线网下载到外部计算机或外部存储设备。网络可以包括铜传输电缆、光纤传输、无线传输、路由器、防火墙、交换机、网关计算机和/或边缘服务器。每个计算/处理设备中的网络适配卡或者网络接口从网络接收计算机可读程序指令，并转发该计算机可读程序指令，以供存储在各个计算/处理设备中的计算机可读存储介质中。
用于执行本公开操作的计算机程序指令可以是汇编指令、指令集架构(ISA)指令、机器指令、机器相关指令、微代码、固件指令、状态设置数据、或者以一种或多种编程语言的任意组合编写的源代码或目标代码，所述编程语言包括面向对象的编程语言—诸如Smalltalk、C++等，以及常规的过程式编程语言—诸如“C”语言或类似的编程语言。计算机可读程序指令可以完全地在用户计算机上执行、部分地在用户计算机上执行、 作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络—包括局域网(LAN)或广域网(WAN)—连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。在一些实施例中，通过利用计算机可读程序指令的状态人员信息来个性化定制电子电路，例如可编程逻辑电路、现场可编程门阵列(FPGA)或可编程逻辑阵列(PLA)，该电子电路可以执行计算机可读程序指令，从而实现本公开的各个方面。
这里参照根据本公开实施例的方法、装置(系统)和计算机程序产品的流程图和/或框图描述了本公开的各个方面。应当理解，流程图和/或框图的每个方框以及流程图和/或框图中各方框的组合，都可以由计算机可读程序指令实现。
这些计算机可读程序指令可以提供给通用计算机、专用计算机或其它可编程数据处理装置的处理器，从而生产出一种机器，使得这些指令在通过计算机或其它可编程数据处理装置的处理器执行时，产生了实现流程图和/或框图中的一个或多个方框中规定的功能/动作的装置。也可以把这些计算机可读程序指令存储在计算机可读存储介质中，这些指令使得计算机、可编程数据处理装置和/或其他设备以特定方式工作，从而，存储有指令的计算机可读介质则包括一个制造品，其包括实现流程图和/或框图中的一个或多个方框中规定的功能/动作的各个方面的指令。
也可以把计算机可读程序指令加载到计算机、其它可编程数据处理装置、或其它设备上，使得在计算机、其它可编程数据处理装置或其它设备上执行一系列操作步骤，以产生计算机实现的过程，从而使得在计算机、其它可编程数据处理装置、或其它设备上执行的指令实现流程图和/或框图中的一个或多个方框中规定的功能/动作。
附图中的流程图和框图显示了根据本公开的多个实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段或指令的一部分，所述模块、程序段或指令的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个连续的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或动作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。
以上已经描述了本公开的各实施例，上述说明是示例性的，并非穷尽性的，并且也不限于所披露的各实施例。在不偏离所说明的各实施例的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。本文中所用术语的选择，旨在最好地解释各实施例的原理、实际应用或对市场中的技术改进，或者使本技术领域的其它普通技术人员能理解本文披露的各实施例。

Claims (21)

  1. 一种信息识别方法，其特征在于，所述方法包括：
    第一设备获取待识别对象的图像序列，所述图像序列包括至少两帧图像；
    所述第一设备从所述图像序列中确定至少一帧图像，作为待识别图像；
    所述第一设备根据所述待识别图像，确定待识别信息；
    所述第一设备向第二设备发送所述待识别信息，以使所述第二设备根据所述待识别信息得到第一识别结果。
  2. 根据权利要求1所述的方法，其特征在于，所述第一设备从所述图像序列中确定至少一帧图像，作为待识别图像，包括：
    所述第一设备根据所述图像序列的识别状态和/或图像质量，从所述图像序列中选定至少一帧图像，作为所述待识别图像。
  3. 根据权利要求2所述的方法，其特征在于，所述第一设备根据所述图像序列的识别状态，从所述图像序列中选定至少一帧图像，作为所述待识别图像，包括：
    所述第一设备对所述图像序列中至少部分图像分别进行第二信息识别，得到一个或多个第二识别结果；
    所述第一设备根据所述一个或多个第二识别结果，得到参考识别结果；
    所述第一设备分别将每个所述第二识别结果与所述参考识别结果进行比较，得到每个所述第二识别结果的置信度；
    所述第一设备根据所述置信度，将全部所述第二识别结果的至少一部分对应的各图像，确定为所述待识别图像，其中，确定出的所述待识别图像的置信度高于所述图像序列中的非待识别图像的置信度。
  4. 根据权利要求2或3所述的方法，其特征在于，所述第一设备根据所述图像序列的图像质量，从所述图像序列中选定至少一帧图像，作为所述待识别图像，包括：
    所述第一设备分别获取所述图像序列中至少部分图像在至少一个衡量维度下的图像质量；
    所述第一设备将所述图像序列中所述图像质量大于对应衡量维度下的阈值的至少一帧图像，确定作为所述待识别图像。
  5. 根据权利要求4所述的方法，其特征在于，所述衡量维度包括清晰度、完整度、强光情况、暗光情况以及遮挡情况中的至少一项。
  6. 根据权利要求1至5中任意一项所述的方法，其特征在于，所述第一设备根据所述待识别图像，确定待识别信息，包括：
    所述第一设备对所述待识别图像进行第一加密处理和/或签名处理，得到待识别信息。
  7. 根据权利要求6所述的方法，其特征在于，所述第一加密处理包括：对所述待识别图像进行编码加密，得到第一加密信息；
    所述签名处理包括：将所述第一设备的签名信息添加至所述待识别图像。
  8. 根据权利要求1至7中任意一项所述的方法，其特征在于，所述待识别对象包 括证卡对象和/或表单对象；
    所述第一识别结果包括所述待识别对象中记录的文本、字符、标识、图标、以及图片中的至少一项。
  9. 一种信息识别方法，其特征在于，所述方法包括：
    第二设备接收待识别对象的待识别信息；
    所述第二设备对所述待识别信息进行第一信息识别，得到第一识别结果；
    所述第二设备根据所述第一识别结果，向第一设备发送识别信息。
  10. 根据权利要求9所述的方法，其特征在于，所述第二设备对所述待识别信息进行第一信息识别，得到第一识别结果，包括：
    所述第二设备获取所述待识别信息包括的待识别图像；
    所述第二设备对所述待识别图像进行防伪检测，得到检测结果；
    在所述检测结果为通过的情况下，所述第二设备对所述待识别图像进行第一信息识别，得到第一识别结果。
  11. 根据权利要求10所述的方法，其特征在于，所述第二设备获取所述待识别信息包括的待识别图像，包括：
    所述第二设备获取所述待识别信息包括的签名信息；
    在获取的所述签名信息与所述第一设备的签名信息匹配的情况下，所述第二设备对所述待识别信息包括的第一加密信息进行解密，得到所述待识别图像。
  12. 根据权利要求9或10所述的方法，其特征在于，所述第二设备对所述待识别图像进行防伪检测，得到检测结果，包括：
    所述第二设备对所述待识别图像进行分类，得到所述待识别图像的分类结果；
    在所述分类结果指示所述待识别图像为通过对所述待识别对象进行拍摄得到的图像的情况下，所述第二设备将检测结果记录为通过；
    和/或，在所述分类结果指示所述待识别图像为通过对所述待识别对象的复印件或翻拍件进行拍摄得到的图像的情况下，所述第二设备将检测结果记录为失败。
  13. 根据权利要求10至12中任意一项所述的方法，其特征在于，所述第二设备对所述待识别信息进行第一信息识别，得到第一识别结果，还包括：
    所述第二设备判断所述第一识别结果是否与预设规则匹配，得到判断结果。
  14. 根据权利要求13所述的方法，其特征在于，所述第二设备根据所述第一识别结果，向第一设备发送识别信息，包括：
    所述第二设备对所述第一识别结果进行第二加密处理，或是对所述第一识别结果和所述判断结果进行所述第二加密处理，得到第二加密信息；
    所述第二设备将所述第二加密信息作为所述识别信息，并向所述第一设备发送。
  15. 根据权利要求10至14中任意一项所述的方法，其特征在于，所述方法还包括：
    所述第二设备保存所述待识别图像中的至少一帧。
  16. 一种信息识别装置，其特征在于，所述装置应用于第一设备并且包括：
    图像序列获取模块，用于获取待识别对象的图像序列，所述图像序列包括至少两帧图像；
    确定模块，用于从所述图像序列中确定至少一帧图像，作为待识别图像；
    待识别信息生成模块，用于根据所述待识别图像，确定待识别信息；
    待识别信息发送模块，用于向第二设备发送所述待识别信息，以使所述第二设备根据所述待识别信息得到第一识别结果。
  17. 一种信息识别装置，其特征在于，所述装置应用于第二设备并且包括：
    接收模块，用于接收待识别对象的待识别信息；
    识别模块，用于对所述待识别信息进行第一信息识别，得到第一识别结果；
    识别信息发送模块，用于根据所述第一识别结果，向第一设备发送识别信息。
  18. 一种电子设备，其特征在于，包括：
    处理器；
    用于存储处理器可执行指令的非暂时性存储介质；
    其中，所述处理器被配置为调用所述存储介质中存储的指令，以执行权利要求1至8中任意一项所述的方法。
  19. 一种电子设备，其特征在于，包括：
    处理器；
    用于存储处理器可执行指令的非暂时性存储介质；
    其中，所述处理器被配置为调用所述存储介质中存储的指令，以执行权利要求9至15中任意一项所述的方法。
  20. 一种计算机可读存储介质，其上存储有计算机程序指令，其特征在于，所述计算机程序指令被处理器执行时实现权利要求1至15任意一项所述的方法。
  21. 一种计算机程序，其中所述计算机程序被处理器执行时，能够实现权利要求1至15任意一项所述的方法。
PCT/CN2020/126167 2020-02-28 2020-11-03 信息识别方法及装置、系统、电子设备、存储介质和计算机程序 WO2021169384A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
KR1020217011452A KR20210110562A (ko) 2020-02-28 2020-11-03 정보 인식 방법, 장치, 시스템, 전자 디바이스, 기록 매체 및 컴퓨터 프로그램
JP2021520536A JP2022524672A (ja) 2020-02-28 2020-11-03 情報認識方法及び装置、システム、電子機器、記憶媒体並びにコンピュータプログラム

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010129842.2A CN111353434A (zh) 2020-02-28 2020-02-28 信息识别方法及装置、系统、电子设备和存储介质
CN202010129842.2 2020-02-28

Publications (1)

Publication Number Publication Date
WO2021169384A1 true WO2021169384A1 (zh) 2021-09-02

Family

ID=71194157

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/126167 WO2021169384A1 (zh) 2020-02-28 2020-11-03 信息识别方法及装置、系统、电子设备、存储介质和计算机程序

Country Status (5)

Country Link
JP (1) JP2022524672A (zh)
KR (1) KR20210110562A (zh)
CN (1) CN111353434A (zh)
TW (1) TWI766458B (zh)
WO (1) WO2021169384A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113792852A (zh) * 2021-09-09 2021-12-14 湖南艾科诺维科技有限公司 基于并行神经网络的信号调制方式识别系统及识别方法
WO2023171623A1 (ja) * 2022-03-07 2023-09-14 ソニーセミコンダクタソリューションズ株式会社 信号処理装置及び信号処理方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111353434A (zh) * 2020-02-28 2020-06-30 北京市商汤科技开发有限公司 信息识别方法及装置、系统、电子设备和存储介质
SG10202006357UA (en) 2020-07-01 2020-09-29 Alipay Labs Singapore Pte Ltd A Document Identification Method and System

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103999442A (zh) * 2011-12-23 2014-08-20 德国捷德有限公司 用于产生数字图像的装置和方法
US20180082121A1 (en) * 2016-09-21 2018-03-22 Cisco Technology, Inc. Method and system for content comparison
CN108229499A (zh) * 2017-10-30 2018-06-29 北京市商汤科技开发有限公司 证件识别方法及装置、电子设备和存储介质
CN110059686A (zh) * 2019-04-26 2019-07-26 腾讯科技（深圳）有限公司 字符识别方法、装置、设备及可读存储介质
CN110688939A (zh) * 2019-09-25 2020-01-14 支付宝(杭州)信息技术有限公司 一种验证待识别证件图像的方法、系统及设备
CN111353434A (zh) * 2020-02-28 2020-06-30 北京市商汤科技开发有限公司 信息识别方法及装置、系统、电子设备和存储介质

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3738631B2 (ja) * 1999-09-27 2006-01-25 三菱電機株式会社 画像検索システムおよび画像検索方法
TWI333089B (en) * 2005-12-30 2010-11-11 Hon Hai Prec Ind Co Ltd Method and system for displaying image focusing
JP2009088944A (ja) * 2007-09-28 2009-04-23 Canon Inc 文字認識装置、撮像装置及び映像再生装置
US8532411B2 (en) * 2009-02-12 2013-09-10 Nippon Telegraph And Telephone Corporation Multi-view image coding method, multi-view image decoding method, multi-view image coding device, multi-view image decoding device, multi-view image coding program, and multi-view image decoding program
CN106302330B (zh) * 2015-05-21 2021-01-05 腾讯科技（深圳）有限公司 身份验证方法、装置和系统
JP6809114B2 (ja) * 2016-10-12 2021-01-06 株式会社リコー 情報処理装置、画像処理システム、プログラム
US11210498B2 (en) * 2017-06-26 2021-12-28 Nec Corporation Facial authentication device, facial authentication method, and program recording medium
JP2019120984A (ja) * 2017-12-28 2019-07-22 トッパン・フォームズ株式会社 帳票データ管理装置、帳票データ管理方法
CN109002842A (zh) * 2018-06-27 2018-12-14 北京字节跳动网络技术有限公司 图像识别方法和装置
CN110796094A (zh) * 2019-10-30 2020-02-14 上海商汤智能科技有限公司 基于图像识别的控制方法及装置、电子设备和存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103999442A (zh) * 2011-12-23 2014-08-20 德国捷德有限公司 用于产生数字图像的装置和方法
US20180082121A1 (en) * 2016-09-21 2018-03-22 Cisco Technology, Inc. Method and system for content comparison
CN108229499A (zh) * 2017-10-30 2018-06-29 北京市商汤科技开发有限公司 证件识别方法及装置、电子设备和存储介质
CN110059686A (zh) * 2019-04-26 2019-07-26 腾讯科技（深圳）有限公司 字符识别方法、装置、设备及可读存储介质
CN110688939A (zh) * 2019-09-25 2020-01-14 支付宝(杭州)信息技术有限公司 一种验证待识别证件图像的方法、系统及设备
CN111353434A (zh) * 2020-02-28 2020-06-30 北京市商汤科技开发有限公司 信息识别方法及装置、系统、电子设备和存储介质

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113792852A (zh) * 2021-09-09 2021-12-14 湖南艾科诺维科技有限公司 基于并行神经网络的信号调制方式识别系统及识别方法
CN113792852B (zh) * 2021-09-09 2024-03-19 湖南艾科诺维科技有限公司 基于并行神经网络的信号调制方式识别系统及识别方法
WO2023171623A1 (ja) * 2022-03-07 2023-09-14 ソニーセミコンダクタソリューションズ株式会社 信号処理装置及び信号処理方法

Also Published As

Publication number Publication date
CN111353434A (zh) 2020-06-30
KR20210110562A (ko) 2021-09-08
TWI766458B (zh) 2022-06-01
TW202133012A (zh) 2021-09-01
JP2022524672A (ja) 2022-05-10

Similar Documents

Publication Publication Date Title
WO2021169384A1 (zh) 信息识别方法及装置、系统、电子设备、存储介质和计算机程序
US11410001B2 (en) Method and apparatus for object authentication using images, electronic device, and storage medium
WO2021017561A1 (zh) 人脸识别方法及装置、电子设备和存储介质
WO2020155828A1 (zh) 图像处理方法及装置、电子设备和存储介质
WO2021031609A1 (zh) 活体检测方法及装置、电子设备和存储介质
WO2020259073A1 (zh) 图像处理方法及装置、电子设备和存储介质
US20210133468A1 (en) Action Recognition Method, Electronic Device, and Storage Medium
TWI702544B (zh) 圖像處理方法、電子設備和電腦可讀儲存介質
WO2022134382A1 (zh) 图像分割方法及装置、电子设备和存储介质、计算机程序
WO2020019760A1 (zh) 活体检测方法、装置及系统、电子设备和存储介质
CN109934275B (zh) 图像处理方法及装置、电子设备和存储介质
WO2017031901A1 (zh) 人脸识别方法、装置及终端
CN105635567A (zh) 拍照方法和装置
WO2021036382A9 (zh) 图像处理方法及装置、电子设备和存储介质
CN104463103B (zh) 图像处理方法及装置
CN110287671B (zh) 验证方法及装置、电子设备和存储介质
CN106296665B (zh) 卡片图像模糊检测方法和装置
TW202038183A (zh) 文本識別方法及裝置、電子設備、儲存介質
WO2022099989A1 (zh) 活体识别、门禁设备控制方法和装置、电子设备和存储介质、计算机程序
WO2022134388A1 (zh) 乘车逃票检测方法及装置、电子设备、存储介质、计算机程序产品
CN110765434A (zh) 身份验证方法、装置、电子设备和存储介质
CN109101542B (zh) 图像识别结果输出方法及装置、电子设备和存储介质
US20220270352A1 (en) Methods, apparatuses, devices, storage media and program products for determining performance parameters
CN107977636B (zh) 人脸检测方法及装置、终端、存储介质
CN109671051A (zh) 图像质量检测模型训练方法及装置、电子设备和存储介质

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2021520536

Country of ref document: JP

Kind code of ref document: A

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20921810

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20921810

Country of ref document: EP

Kind code of ref document: A1