CN111639970A

CN111639970A - Method for determining price of article based on image recognition and related equipment

Info

Publication number: CN111639970A
Application number: CN202010468311.6A
Authority: CN
Inventors: 高立志
Original assignee: OneConnect Financial Technology Co Ltd Shanghai
Current assignee: OneConnect Smart Technology Co Ltd; OneConnect Financial Technology Co Ltd Shanghai
Priority date: 2020-05-28
Filing date: 2020-05-28
Publication date: 2020-09-08

Abstract

The application relates to the field of artificial intelligence, and provides an article price determining method based on image recognition and related equipment. The method comprises the following steps: acquiring a target image including a target object; identifying the target image to obtain a plurality of characteristic information representing the identity of the target object; performing relevance analysis on the plurality of feature information, and determining target feature information in the plurality of feature information based on the analysis result; and analyzing the target characteristic information through linear regression to determine the price of the target object. The implementation of this application is favorable to reducing the time cost of carrying out price estimation to article, improves the efficiency and the accuracy of price estimation.

Description

Method for determining price of article based on image recognition and related equipment

Technical Field

The application relates to the field of artificial intelligence, in particular to an article price determining method based on image recognition and related equipment.

Background

Price evaluation is a pre-procedure in many areas of life, such as item auction, item recycling, etc., and during item auction, before the item is auctioned, the item is generally evaluated in price by an asset evaluation organization with corresponding qualification.

In the prior art, when evaluating the price of an auctioned item, the auction method generally depends on a manual review system of an asset evaluation organization and information related to the auctioned item. The manual review method consumes more time when collecting the related information of the auction articles and evaluating the articles, has low efficiency and high labor cost, and is difficult to avoid price errors caused by personal subjectivity or calculation errors in the review process.

Disclosure of Invention

This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

In a first aspect of the present application, there is provided an article price determining method based on image recognition, including: acquiring a target image including a target object; identifying the target image to obtain a plurality of characteristic information representing the identity of the target object; performing relevance analysis on the plurality of feature information, and determining target feature information in the plurality of feature information based on the analysis result; and analyzing the target characteristic information through linear regression to determine the price of the target object.

With reference to the first aspect, in a first implementation manner of the first aspect, the identifying the target image and obtaining a plurality of feature information characterizing an identity of the target object includes: respectively calculating first similarity of the target image and each reference image in a database; extracting a first target area image where a preset marker is located in the target image and a second target area image where the preset marker is located in each reference image; calculating a second similarity of the first target area image and the second target area image; weighting and calculating the first similarity and the second similarity to determine comprehensive similarity; determining at least one reference image corresponding to the comprehensive similarity higher than a first preset threshold as a similar image, and taking a plurality of feature information of an article corresponding to the similar image as a plurality of feature information representing the identity of the target object; the plurality of characteristic information includes at least one of a brand, a model, a production date, an original price, and a transaction price.

With reference to the first aspect, in a second implementation manner of the first aspect, the identifying the target image and obtaining a plurality of feature information characterizing an identity of the target object includes: respectively calculating first similarity of the target image and each reference image in a database; determining a reference article corresponding to the reference images with the highest first similarity and the preset values; determining depreciation categories corresponding to the preset numerical value reference articles, and taking the depreciation category with the largest number of the corresponding reference articles as the depreciation category corresponding to the target object; and determining characteristic information which represents the identity of the target object and is associated with the depreciation rate based on the depreciation category corresponding to the target object and a preset depreciation coefficient.

With reference to the first aspect, in a third implementation of the first aspect, the identifying the target image and obtaining a plurality of attribute feature information characterizing an identity of the target object includes: matching the plurality of characteristic information with a preset characteristic list, and determining the information grade of each characteristic information; the preset feature list comprises feature information and information levels with mapping relations; and determining at least one piece of characteristic information in the plurality of pieces of characteristic information based on the information grade, inputting the characteristic information into a preset depreciation model, and obtaining the characteristic information which represents the identity of the target object and is associated with the depreciation rate.

With reference to the first aspect, in a fourth implementation manner of the first aspect, the performing a correlation analysis on the plurality of feature information, and determining target feature information in the plurality of feature information based on a result of the analysis includes: performing relevance analysis on the plurality of characteristic information; determining at least two pieces of feature information to be related, excluding binary features in the at least two pieces of feature information, and fusing the at least two pieces of feature information with the related coefficient higher than a second preset threshold; and forming target characteristic information based on the plurality of characteristic information after the elimination and/or fusion processing.

With reference to the fourth implementation manner of the first aspect, in a fifth implementation manner of the first aspect, after the forming target feature information based on the plurality of feature information after the excluding and/or fusing processes, the forming the target feature information includes: acquiring a plurality of pieces of feature information after elimination and/or fusion processing, and mapping each piece of feature information to a corresponding preset feature tag; and determining at least one preset feature tag to have no feature information corresponding to the preset feature tag, and/or determining at least one feature information corresponding to a target feature tag in the preset feature tags, and reporting the target feature information.

With reference to the fourth implementation manner or the fifth implementation manner of the first aspect, in a sixth implementation manner of the first aspect, the determining the price of the target object by analyzing the target feature information through linear regression includes: performing the step of calculating the target object price using a linear regression model trained to converge: determining corresponding linear regression coefficients based on the target characteristic information, and constructing a corresponding linear regression equation; calculating the price of the target object based on the target characteristic information and a linear regression equation; wherein the training of the linear regression model comprises: training a training set input model to obtain a plurality of groups of training coefficients for constructing a linear regression equation; inputting target characteristic information corresponding to an article with a known real price into a linear regression model to obtain a predicted price output by the model; comparing the predicted price with the real price, and if the price error is within a preset range, determining the training coefficient as a linear regression coefficient; the training set comprises target characteristic information corresponding to a plurality of groups of articles with known real prices and real prices thereof.

In a second aspect of the present application, there is provided an article price determining apparatus based on image recognition, comprising: an acquisition module for acquiring a target image including a target object; the identification module is used for identifying the target image and acquiring a plurality of characteristic information representing the identity of the target object; the analysis module is used for carrying out correlation analysis on the characteristic information and determining target characteristic information in the characteristic information based on the analysis result; and the determining module is used for determining the price of the target object by analyzing the target characteristic information through linear regression.

With reference to the second aspect, in a first implementation manner of the second aspect, the identification module includes: the calculating unit is used for respectively calculating first similarity of the target image and each reference image in the database; the extraction unit is used for extracting a first target area image where a preset marker is located in the target image and a second target area image where the preset marker is located in each reference image; calculating a second similarity of the first target area image and the second target area image; the weighting unit is used for weighting and calculating the first similarity and the second similarity to determine comprehensive similarity; a first determining unit, configured to determine that at least one reference image corresponding to the comprehensive similarity higher than a first preset threshold is used as a similar image, and use a plurality of feature information of an article corresponding to the similar image as a plurality of feature information representing the identity of the target object; the plurality of characteristic information includes at least one of a brand, a model, a production date, an original price, and a transaction price.

With reference to the second aspect, in a second implementation manner of the second aspect, the identification module includes: the calculating unit is used for respectively calculating first similarity of the target image and each reference image in the database; a reference article determining unit, configured to determine a reference article corresponding to a number of reference images with a preset value, where the first similarity is highest; the depreciation category determining unit is used for determining depreciation categories corresponding to the preset numerical value reference articles, and taking the depreciation category with the largest number of the corresponding reference articles as the depreciation category corresponding to the target object; and the second determining unit is used for determining characteristic information which represents the identity of the target object and is associated with the depreciation rate based on the depreciation category corresponding to the target object and a preset depreciation coefficient.

With reference to the second aspect, in a third implementation of the second aspect, the identification module includes: the matching unit is used for matching the plurality of characteristic information with a preset characteristic list and determining the information grade of each characteristic information; the preset feature list comprises feature information and information levels with mapping relations; and the third determining unit is used for determining at least one piece of feature information in the plurality of pieces of feature information based on the information levels, inputting the feature information into a preset depreciation model, and obtaining feature information which represents the identity of the target object and is associated with the depreciation rate.

With reference to the second aspect, in a fourth implementation of the second aspect, the analysis module includes: the analysis unit is used for carrying out correlation analysis on the characteristic information; the processing unit is used for determining that at least two pieces of feature information are related, eliminating binary features in the at least two pieces of feature information, and fusing at least two pieces of feature information with the related coefficient higher than a second preset threshold value; and the generating unit is used for forming target characteristic information based on the plurality of characteristic information after the elimination and/or fusion processing.

With reference to the fourth implementation manner of the second aspect, in a fifth implementation manner of the second aspect, the analysis module further includes: the mapping unit is used for acquiring a plurality of pieces of feature information after elimination and/or fusion processing, and mapping each piece of feature information to a corresponding preset feature label; and the reporting unit is used for reporting the target characteristic information if at least one preset characteristic label is determined to have no characteristic information corresponding to the preset characteristic label and/or at least one characteristic information corresponding to a target characteristic label in the preset characteristic labels.

With reference to the fourth or fifth implementation manner of the second aspect, in a sixth implementation manner of the second aspect, the determining module includes: an execution unit for performing the step of calculating the target object price using a linear regression model trained to converge: determining corresponding linear regression coefficients based on the target characteristic information, and constructing a corresponding linear regression equation; calculating the price of the target object based on the target characteristic information and a linear regression equation; wherein the training of the linear regression model comprises: training a training set input model to obtain a plurality of groups of training coefficients for constructing a linear regression equation; inputting target characteristic information corresponding to an article with a known real price into a linear regression model to obtain a predicted price output by the model; comparing the predicted price with the real price, and if the price error is within a preset range, determining the training coefficient as a linear regression coefficient; the training set comprises target characteristic information corresponding to a plurality of groups of articles with known real prices and real prices thereof.

In a third aspect of the present application, an electronic device is provided, including: a memory and a processor; the memory has a computer program stored therein; a processor for performing the method of the first aspect and any of its embodiments when running a computer program.

In a fourth aspect of the present application, a computer-readable medium is provided, on which a computer program is stored, which computer program, when being executed by a processor, is adapted to carry out the method of the first aspect and any of its embodiments.

The beneficial effect that technical scheme that this application provided brought is:

the method comprises the steps of obtaining a target image comprising a target object; identifying the target image based on an image identification technology to obtain a plurality of characteristic information representing the identity of the target object; performing relevance analysis on the plurality of feature information, and determining target feature information in the plurality of feature information based on the analysis result; and analyzing the target characteristic information through linear regression to determine the price of the target object. The implementation of this application is favorable to reducing the time cost of carrying out price estimation to article, improves the efficiency and the accuracy of price estimation.

Drawings

The above and other features, advantages and aspects of various embodiments of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numbers refer to the same or similar elements. It should be understood that the drawings are schematic and that elements and features are not necessarily drawn to scale.

Fig. 1 is a flowchart of an item price determination method based on image recognition according to an embodiment of the present application;

FIG. 2 is a flowchart of an item price determination method based on image recognition according to an embodiment of the present application;

FIG. 3 is a flowchart of an item price determination method based on image recognition according to an embodiment of the present application;

FIG. 4 is a flowchart of an item price determination method based on image recognition according to an embodiment of the present application;

FIG. 5 is a flowchart of an item price determination method based on image recognition according to an embodiment of the present application;

FIG. 6 is a flowchart of an item price determination method based on image recognition according to an embodiment of the present application;

FIG. 7 is a flowchart of an item price determination method based on image recognition according to an embodiment of the present application;

fig. 8 is a schematic structural diagram of an article price determining apparatus based on image recognition according to an embodiment of the present application;

fig. 9 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

Embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but rather are provided for a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are for illustration purposes only and are not intended to limit the scope of the present application.

It should be understood that the various steps recited in the method embodiments of the present application may be performed in a different order and/or in parallel. Moreover, method embodiments may include additional steps and/or omit performing the illustrated steps. The scope of the present application is not limited in this respect.

The term "include" and variations thereof as used herein are open-ended, i.e., "including but not limited to". The term "based on" is "based, at least in part, on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions for other terms will be given in the following description.

It should be noted that the terms "first", "second", and the like in the present application are only used for distinguishing the devices, modules or units, and are not used for limiting the devices, modules or units to be different devices, modules or units, and are not used for limiting the sequence or interdependence relationship of the functions executed by the devices, modules or units.

It is noted that references to "a", "an", and "the" modifications in this application are intended to be illustrative rather than limiting, and that those skilled in the art will recognize that reference to "one or more" unless the context clearly dictates otherwise.

The names of messages or information exchanged between a plurality of devices in the embodiments of the present application are for illustrative purposes only, and are not intended to limit the scope of the messages or information.

To make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings.

The terms referred to in this application will first be introduced and explained:

and (3) correlation analysis: the correlation analysis is used for analyzing the correlation among a plurality of characteristic information, such as the characteristic information comprises: analyzing the correlation between the brand and the original price, the brand and the transaction price, and the original price and the transaction price; specifically, the correlation analysis can be performed by SPSS. In the result of the correlation analysis, firstly, judging whether the two pieces of characteristic information have a relationship; when the two have a relationship, judging the relationship to be positive correlation or negative correlation; then, the degree of closeness of the relationship between the two is judged, and the closeness is generally expressed by a correlation coefficient.

Target characteristic information: the target characteristic information is composed of a plurality of characteristic information; the characteristic information is obtained by identifying a target image comprising a target object, and a plurality of characteristic information can be obtained by image detection through an OCR technology picture contrast technology, multi-scale characteristic fusion of image detection, or image identification through an image identification model and the like. After performing relevance analysis on the plurality of feature information, determining target feature information in the plurality of feature information based on the analysis result; the target feature information may be one or more of a plurality of feature information, or may be generated by fusing one feature information formed by a plurality of feature information.

The following describes the technical solutions of the present application and how to solve the above technical problems with specific embodiments. The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments. Embodiments of the present application will be described below with reference to the accompanying drawings.

Referring to fig. 1, the present application provides an article price determining method based on image recognition, which may be specifically executed by an electronic device according to an embodiment of the present application, and may be specifically a server or a terminal device; when the system of the application is built, the database of the system comprises images of a plurality of articles and corresponding characteristic information thereof, such as the brand, model, depreciation category, production date and the like of the articles. In one embodiment, the present application may be applied to the field of estimation of property value of a court performer, where the database includes property information of the performer over years, and the present embodiment is described as being applied to the field of estimation of property price of a court performer. The method comprises the following steps:

s101, acquiring a target image comprising a target object;

s102, identifying the target image and obtaining a plurality of characteristic information representing the identity of the target object;

s103, performing relevance analysis on the plurality of characteristic information, and determining target characteristic information in the plurality of characteristic information based on the analysis result;

s104, analyzing the target characteristic information through linear regression to determine the price of the target object.

Specifically, in step S101, a target image including a target object is acquired, which may be the target image input by the user, or the target image including the target object is determined based on the image input by the user (i.e., before step S101, the image input by the user is detected). The target object includes various entities such as house-like buildings, machine equipment, transportation equipment, tool and tools, living goods, and the like. If the target object is a vehicle, step S101 obtains a target image including the vehicle. The target image is specific to the current state of the article for which the price evaluation of the article is performed, such as a picture, a video and the like which are instantly taken based on the article.

Specifically, in step S102, the target image is identified, and a plurality of feature information characterizing the identity of the target object is obtained; the recognition of the target image may be performed by an OCR technology image comparison technology, or by other image recognition technologies, and the image recognition may be completed by applying the prior art, which is not repeated herein. In the image recognition process, a plurality of characteristic information which embodies the image content can be extracted. When the target image of the vehicle is identified, the characteristic information including the brand name, the license plate number and the like of the vehicle can be identified.

Specifically, in step S103, performing correlation analysis on the plurality of feature information, and determining target feature information among the plurality of feature information based on the analysis result; when the correlation analysis is performed, two pieces of feature information are generally performed, and when the feature information representing the identity of the target object obtained in step S102 includes three or more pieces of feature information, such as feature information a, feature information B, and feature information C, the three pieces of feature information are analyzed one by one (a and B, A and C, and B and C are analyzed). Wherein, the analysis result of the correlation analysis comprises: the feature information is irrelevant, the feature information is positively correlated, the feature information is negatively correlated, and the feature information is correlated with each other (the correlation coefficient is larger, the correlation between the two features is higher). The target feature information is determined from the plurality of feature information based on the correlation analysis result, and may be one or more of the plurality of feature information and/or a combination of the plurality of feature information. If a plurality of characteristic information comprises A, B, C, D, wherein A is not related to B, C, D, B is positively related to C, and the correlation coefficient of C and D is close to 0.8 (the characteristic correlation is high), the C and D are fused to form characteristic information F; the target characteristic information includes A, B and F. The fusion of the characteristic information can be normalized by adopting a weighting calculation method.

Specifically, in step S104, analyzing the target feature information through linear regression to determine the price of the target object; in the linear regression calculation, target characteristic information is an independent variable, and the price of a target object is a dependent variable; in the calculation, the target feature information may include one piece of feature information, or may include a plurality of pieces of feature information, and the linear regression calculation is performed with different linear regression coefficients (for example, in y ═ ax + b, a and b are linear regression coefficients, x is an independent variable, and y is a dependent variable) based on the difference in the target feature information, and when the target feature information includes a plurality of pieces of feature information, y ═ a1x1+ a2x2+ … aNxN + b). Further, the step of step S104 may be completed by a linear regression model, and the target feature information obtained in step S103 is input into the linear regression model to obtain the price of the target object.

In an embodiment, the four steps of steps S101 to S104 may be formed by building a composite model, for example, building a price evaluation model, where the model includes an image recognition module, a correlation analysis module, and a linear regression module; inputting a target image including a target object into a model; carrying out image recognition through an image recognition module to obtain a plurality of characteristic information; transferring the plurality of characteristic information to a correlation analysis module for correlation analysis to obtain target characteristic information; and transferring the target characteristic information to a linear regression module for price calculation, and obtaining and outputting the price of the target object.

In one embodiment, the target image including the target object acquired in step S101 includes a receipt for purchasing the item, an instruction for using the item, and the like, in addition to the image photographed based on the item; by identifying the target image in step S102, characteristic information such as a brand name, a model number, a purchase price, and the like related to the article in the document can be identified. At this time, the target image including the target object in step S101 is not limited to the target image including the form expression of the target object, but also includes a character expression, and if the target image is a receipt of a purchase item, the receipt may have content representing the target object, which may be characters, symbols, and the like.

The embodiment obtains a target image including a target object; identifying the target image based on an image identification technology to obtain a plurality of characteristic information representing the identity of the target object; performing relevance analysis on the plurality of feature information, and determining target feature information in the plurality of feature information based on the analysis result; and analyzing the target characteristic information through linear regression to determine the price of the target object. The implementation of this application is favorable to reducing the time cost of carrying out price estimation to article, improves the efficiency and the accuracy of price estimation.

In an embodiment, referring to fig. 2, the step S102 of recognizing the target image and obtaining a plurality of feature information characterizing the identity of the target object includes:

s201, respectively calculating first similarity of the target image and each reference image in a database;

s202, extracting a first target area image where a preset marker is located in the target image and a second target area image where the preset marker is located in each reference image; calculating a second similarity of the first target area image and the second target area image;

s203, weighting and calculating the first similarity and the second similarity to determine comprehensive similarity;

s204, determining at least one reference image corresponding to the comprehensive similarity higher than a first preset threshold as a similar image, and taking a plurality of feature information of an article corresponding to the similar image as a plurality of feature information representing the identity of the target object; the plurality of characteristic information includes at least one of a brand, a model, a production date, an original price, and a transaction price.

The calculation of the first similarity and the second similarity is to calculate the image similarity of the two images, and is mainly used for scoring the similarity of contents between the two images and judging the similarity of the contents of the images according to the degree of the score; the similarity calculation can be carried out by adopting methods such as histogram matching, cosine similarity calculation and the like.

Specifically, in step S201, first similarities between the target image and each reference image in the database are respectively calculated; wherein, the database comprises information of property of the executed person for many years, such as the corresponding brand, model, depreciation category, exchange rate, production date, original price, transaction price and the like of vehicles, houses and luxury goods; further, the database also includes images of property items, such as images of original purchases of vehicles, documents when purchased of vehicles, etc.; besides forming a database by the information of the executed person about the property, the database can be enriched by common information, such as different brands of vehicles, models, attributes, prices and the like corresponding to the different brands. Taking an image included in the database as a reference image, and calculating a first similarity between the target image and the reference image; since the database includes a plurality of reference images, a first similarity of the target image compared with each reference image is calculated in step S201; that is, when the database includes A, B and C reference images, the first similarity between the target image and the reference image a, the first similarity between the target image and the reference image B, and the first similarity between the target image and the reference image C are calculated.

In step S202, a first target area image in which a preset marker is located in the target image and a second target area image in which the preset marker is located in each reference image are extracted; calculating a second similarity of the first target area image and the second target area image; when the target image is a vehicle, the preset marker can be an automobile marker logo. When the first target area image and the second target area image are extracted, the area where the preset marker is located in the target image can be detected through an image detection technology, the area where the preset marker is located can be determined through a rectangle, a circle, an ellipse, an irregular figure or the like, the area where the preset marker is located in the target image is divided into the first target area image, and the area where the preset marker is located in the reference image is divided into the second target area image. In one embodiment, the preset identifier is set based on the target image, and there may not be a second target area image where the preset identifier is located in the reference image, and thus, the method includes at least two processing methods: (1) taking an image area which is closest to a preset marker in the reference image as a second target area image; (2) and setting the second target area image in which the preset identifier in the reference image is located as null, wherein the second similarity value between the first target area image and the second target area image is 0. With respect to the first similarity calculated in step S201, step S202 calculates a second similarity for the first target area image and the second target area image to further determine the similarity of the target image and each reference image.

In step S203, the first similarity and the second similarity are weighted and calculated to determine a comprehensive similarity; in one embodiment, since the second similarity is calculated based on the first target area image and the second target area image, and the content expressed by the second similarity is limited, the weight occupied by the second similarity may be set to be lower than the first similarity; if the weight occupied by the first similarity is 0.7, the weight occupied by the second similarity is 0.3; the overall similarity is determined by a weighted calculation, such as the overall similarity is 0.7+ 0.3.

In step S204, determining at least one reference image corresponding to the integrated similarity higher than a first preset threshold as a similar image, and taking a plurality of feature information of an article corresponding to the similar image as a plurality of feature information characterizing the identity of the target object; the plurality of characteristic information includes at least one of a brand, a model, a production date, an original price, and a transaction price; the first preset threshold value can be adjusted based on the calculation result of the comprehensive similarity, and if the comprehensive similarity is generally low, the first preset threshold value is reduced to obtain a reference image with the highest comprehensive similarity as a similar image; if the comprehensive similarity is generally high, increasing a first preset threshold value to obtain at least one reference image with the highest comprehensive similarity as a similar image; in general, the first preset threshold is a default value, for example, set to 0.7, and if the overall similarity between 3 reference images and the target image is higher than 0.7, all the 3 reference images are regarded as similar images, and a plurality of feature information of an article corresponding to each of the 3 reference images is regarded as a plurality of feature information representing the identity of the target object, that is, the plurality of feature information representing the identity of the target object is constituted by the feature information of the article corresponding to each of the 3 reference images.

Further, step S202 extracts a first target area image in which a preset marker is located in the target image, and a second target area image in which the preset marker is located in each reference image; calculating a second similarity between the first target area image and the second target area image, including: acquiring at least one reference image with the first similarity higher than a third preset threshold; extracting a first target area image where a preset marker is located in the target image, and extracting a second target area image where the preset marker is located in at least one reference image with the first similarity higher than a third preset threshold; and calculating a second similarity of the first target area image and the second target area image.

Specifically, since the database includes a plurality of reference images, in order to improve the accuracy of image recognition and reduce the complexity of calculation, after the first similarity is obtained through calculation, at least one reference image with the first similarity higher than a third preset threshold is obtained, and the second similarity calculation is performed on the target area image with respect to the target image. If the database comprises reference images a, b and c, wherein the first similarity between the reference images a and b and the target image is higher than a third preset threshold value, second target area images where preset markers in the reference images a and b are located are respectively extracted, and second similarity calculation is carried out between the second target area images and the first target area images where the preset markers in the target image are located.

In an embodiment, referring to fig. 3, the step S102 of recognizing the target image and obtaining a plurality of feature information characterizing the identity of the target object includes:

s301, respectively calculating first similarity of the target image and each reference image in a database;

s302, determining a reference article corresponding to the reference images with the highest first similarity and the preset values;

s303, determining the depreciation categories corresponding to the preset numerical value reference articles, and taking the depreciation category with the largest number corresponding to the reference articles as the depreciation category corresponding to the target object;

s304, determining characteristic information which represents the identity of the target object and is associated with the depreciation rate based on the depreciation category corresponding to the target object and a preset depreciation coefficient.

The results of step S301 and step S201 are the same, and are not described herein again. Specifically, after the first similarity between each reference image and the target image is determined through calculation, the reference articles corresponding to the first preset number of reference images with the highest first similarity are determined (for example, the reference articles corresponding to the first 5 reference images with the highest first similarity are determined), based on the depreciation categories corresponding to the preset number of reference articles, the depreciation category with the largest number of corresponding reference articles is taken as the depreciation category corresponding to the target object (for example, the depreciation categories include five types, namely new, newer, medium, older, old, and the like, the reference articles corresponding to the 5 reference images respectively correspond to the depreciation categories of the reference articles, and if the depreciation category is medium and corresponds to 3 reference articles, the depreciation category is medium and is taken as the depreciation category of the target object); in step S304, determining a corresponding depreciation coefficient based on the depreciation category of the target object determined in step S303, calculating a depreciation rate, where the calculated depreciation rate is feature information representing the identity of the target object and associated with the depreciation rate; in the embodiment, a KNN algorithm can be adopted, for example, a classification model is built, five marked (new, newer, medium, older and old) depreciation categories are counted through mass data, and a depreciation coefficient corresponding to each depreciation category is determined; after the target image is input into the classification model, the first K reference images (reference images with preset values) most similar to the target image are determined, the depreciation category corresponding to the target image is the depreciation category with the largest occurrence frequency in the depreciation categories corresponding to the first K reference images, the target image is determined to be one of the five marked depreciation categories, so that the depreciation category of the target image is determined, and the depreciation rate of the target object is determined according to the depreciation coefficient corresponding to the depreciation category.

In an embodiment, referring to fig. 4, the step S102 of recognizing the target image and obtaining a plurality of feature information characterizing the identity of the target object includes:

s401, matching the plurality of feature information with a preset feature list, and determining the information grade of each feature information; the preset feature list comprises feature information and information levels with mapping relations;

s402, determining at least one piece of feature information in the feature information based on the information grade, inputting the feature information into a preset depreciation model, and obtaining feature information which represents the identity of the target object and is associated with the depreciation rate.

The preset feature list is described with reference to table 1:

TABLE 1

As can be seen from table 1, the characteristic information corresponding to the year of production, the original price, etc. corresponds to the first rank; the feature information, which is or is not entirely new, corresponds to null; it is assumed that the plurality of feature information for characterizing the identity of the target object in step S401 includes: determining the production date and the original price as a first grade if the production date, the original price, the brand new and the depreciation rate are the same; determining to be empty null; the depreciation rate is determined as the nth level. In step S402, determining at least one piece of feature information from the plurality of pieces of feature information based on the information level, and inputting the feature information into a preset depreciation model to obtain feature information representing the identity of the target object and associated with a depreciation rate; wherein, because the production date and the original price both correspond to the first grade, choose a characteristic information in the first grade at random, such as choosing the production date; deleting the characteristic information because the brand new corresponding information level is null; since only the depreciation rate corresponds to the Nth level, the depreciation rate is also the selected characteristic information; inputting preset characteristic information of the depreciation model at the moment, wherein the characteristic information comprises a production date and a depreciation rate; the preset depreciation model may refer to the classification model constructed based on the KNN algorithm in steps S301 to S304, and the difference is that the input data of the depreciation model of this embodiment is feature information, and the preset depreciation model determines the depreciation rate of the target object through analysis processing of the feature information.

In an embodiment, referring to fig. 5, in step S103, performing a correlation analysis on the feature information, and determining target feature information in the feature information based on the analysis result, including:

s501, performing relevance analysis on the characteristic information;

s502, determining that at least two pieces of feature information are related, excluding binary features in the at least two pieces of feature information, and fusing at least two pieces of feature information with a correlation coefficient higher than a second preset threshold;

s503 forms target feature information based on the plurality of feature information after the exclusion and/or fusion process.

In step S501, performing correlation analysis on the plurality of feature information; specifically, when performing the correlation analysis, two feature information are compared, and if a plurality of feature information such as the brand, the model, the production date, the original price, the brand new, the transaction price, and the depreciation rate are obtained in step S102, then the correlation analysis is performed on the basis of each feature information and other feature information, such as the brand is analyzed with the model, the production date, the original price, the brand new, the transaction price, and the depreciation rate, and the model is analyzed with the brand, the production date, the original price, the brand new, the transaction price, and the depreciation rate, and so on. And performing correlation analysis to determine whether the two feature information are correlated, if so, determining positive correlation or negative correlation, and further determining a correlation coefficient (correlation compactness) of the two feature information.

In step S502, it is determined that at least two pieces of feature information are related, a binary feature in the at least two pieces of feature information is excluded, and at least two pieces of feature information having a correlation coefficient higher than a second preset threshold are fused; the binary feature represents forward information or backward information of the feature information, for example: whether the feature information is brand new, the binary feature is only characterized as yes or no. When the two pieces of feature information are determined to be related, firstly removing the binary features, wherein if the depreciation rate and brand-new feature information are related, if the brand-new feature information is the binary feature, the brand-new feature is removed; if the correlation coefficient of the depreciation rate and other feature information is higher than a second preset threshold value after the brand-new condition is eliminated, the depreciation rate and the feature information are fused to form feature information, and the fusion can be normalized in a weighting calculation mode. For example, the following steps are carried out: the feature information for performing the correlation analysis is assumed to include: brand, model, date of manufacture, original price, brand new, trade price and depreciation rate; wherein, the depreciation rate is related to brand new, and the brand new is a binary characteristic; wherein the production date is related to the original price, and the correlation coefficient is higher than a second preset threshold value; the plurality of feature information after the elimination and/or fusion processing in step S502 includes: brand, model, transaction price, depreciation rate, and a piece of feature information (fused feature information) formed by fusing the date of manufacture with the original price. Further, in step S503, the brand, model, transaction price, depreciation rate, and one piece of feature information (fused feature information) formed by fusing the production date and the original price are set as the target feature information.

In one implementation, referring to fig. 6, after the step S503 forms the target feature information based on the plurality of feature information after the excluding and/or fusing processes, the method further includes:

s601, acquiring a plurality of pieces of feature information after elimination and/or fusion processing, and mapping each piece of feature information to a corresponding preset feature tag;

s602, if at least one preset feature tag is determined to have no feature information corresponding to the preset feature tag, or at least one feature information corresponds to a target feature tag in the preset feature tags, the target feature information is reported.

In step S601, a plurality of pieces of feature information after elimination and/or fusion processing are obtained, and each piece of feature information is mapped to a corresponding preset feature tag; the preset feature tag is used for identifying feature information with a large influence in calculating the price of an article, and if a target object is a house, feature information representing the size of the house (the preset feature tag corresponding to the size of the article) is lacked; or if the target object is blue and white porcelain, the price which is correspondingly calculated as a collection (target characteristic label corresponding to the characteristic information) is higher than the original price to a certain extent.

In step S602, if it is determined that at least one preset feature tag has no feature information corresponding thereto, and/or at least one feature information corresponds to a target feature tag in the preset feature tags, reporting the target feature information; specifically, the description is made with reference to the following table 2:

TABLE 2

Object feature information	Preset feature tag	Object feature tag
				Size of article
23 years of collection	Collection	*
			Year of production 1990	Production time
Depreciation rate 0.6	Depreciation rate
			…	…

As can be seen from table 2 above, the preset feature tag represents a feature information type having a large influence in calculating the price of the target object in this embodiment. The target feature information comprises 23 years of collection, and is mapped to a column of a collection with preset feature tags (the collection is the target feature tag); and if the preset feature tag 'article size' does not have any feature information corresponding to the object size, determining that one preset feature tag has no feature information corresponding to the object size, and reporting the target feature information because the target feature information 'collection year 23 years' corresponds to the target feature tag 'collection article' in the preset feature tags. In an embodiment, reporting the target feature information may be understood as sending an alarm, reporting the target feature information to the server for information storage, and drawing the user' S attention to whether the price calculated in step S104 needs to be adjusted. Further, the reported target feature information may also be filled in by the server, where if the size of the currently missing article is determined, the server queries the relevant information of the target object in the database, and if the size of the article related to the target object can be obtained, the feature information corresponding to the preset feature tag "article size" is filled in step S602.

In an embodiment, referring to fig. 7, the step S104 of determining the price of the target object by analyzing the target feature information through linear regression includes:

s701, the step of calculating the target object price by adopting a linear regression model trained to be converged is executed: determining corresponding linear regression coefficients based on the target characteristic information, and constructing a corresponding linear regression equation; calculating the price of the target object based on the target characteristic information and a linear regression equation;

wherein the training of the linear regression model comprises: training a training set input model to obtain a plurality of groups of training coefficients for constructing a linear regression equation; inputting target characteristic information corresponding to an article with a known real price into a linear regression model to obtain a predicted price output by the model; comparing the predicted price with the real price, and if the price error is within a preset range, determining the training coefficient as a linear regression coefficient; the training set comprises target characteristic information corresponding to a plurality of groups of articles with known real prices and real prices thereof.

Specifically, the linear regression model built by the embodiment can adapt to various expressions of target characteristic information, for example, the target characteristic information includes one or more pieces of characteristic information; when the target feature information only comprises one feature information 'original price', corresponding linear regression coefficients a and b are determined, and a corresponding linear regression equation (origin _ price) ═ a _ origin _ price + b is constructed, wherein f (x) represents a calculation formula of a target value. When the target feature information comprises feature information of 'date, quantity and original price', determining corresponding linear regression coefficients a1, a2, a3 and b, and constructing a corresponding linear regression equation price ═ f (date, number, origin _ price) ═ a1 ═ date + a2 × number + a3 origin _ price + b; and after the linear regression equation is constructed, calculating to obtain the price of the target object.

The linear regression coefficients can be obtained during model training, for example, a training set is input into the linear regression model to train multiple groups of training coefficients for constructing multiple linear regression equations, wherein the training set comprises target feature information corresponding to multiple groups of articles with known real prices and real prices of the articles; when the model selects the optimal linear regression coefficient, inputting target characteristic information corresponding to the article with the known real price into the linear regression model to obtain the predicted price output by the model; and comparing the predicted price with a real price (E ═ error ═ price-predicted _ price |, wherein price is the real price, predicted _ price is the predicted price, and error is the price error), and if the price error is within a preset range, determining that the training coefficient is a linear regression coefficient.

In one embodiment, the method for determining the price of an item based on image recognition is applied to a scene of auction of the property of an executed person in the process of executing a court, and the database comprises evaluation information of the property of the executed person and images of related properties. When the house of the executed person A needs to be auctioned, the price of the house needs to be estimated firstly, the executed person or law enforcement personnel can input a two-dimensional plane graph, a three-dimensional stereo graph, a house purchase contract image or a field shot effect graph and the like representing the house into equipment, the equipment transmits the acquired image to a server, the server identifies the image to obtain a plurality of pieces of characteristic information corresponding to the house, correlation analysis is carried out on the image based on the plurality of pieces of characteristic information to determine target characteristic information, then the target characteristic information is analyzed by linear regression to determine the current price of the house and fed back to the equipment, and finally the estimated price of the house is displayed on a display interface of the equipment.

Referring to fig. 8, a schematic structural diagram of an article price determining apparatus based on image recognition according to an embodiment of the present application is provided, and an article price determining apparatus 800 based on image recognition according to an embodiment of the present application may include:

an acquisition module 801 for acquiring a target image including a target object; the identification module 802 is configured to identify the target image and obtain a plurality of feature information representing the identity of the target object; an analysis module 803, configured to perform correlation analysis on the plurality of feature information, and determine target feature information in the plurality of feature information based on an analysis result; a determining module 804, configured to determine the price of the target object by analyzing the target feature information through linear regression.

In one embodiment, the identification module 802 includes: the calculating unit is used for respectively calculating first similarity of the target image and each reference image in the database; the extraction unit is used for extracting a first target area image where a preset marker is located in the target image and a second target area image where the preset marker is located in each reference image; calculating a second similarity of the first target area image and a second target area image; the weighting unit is used for weighting and calculating the first similarity and the second similarity to determine comprehensive similarity; a first determining unit, configured to determine that at least one reference image corresponding to the comprehensive similarity higher than a first preset threshold is used as a similar image, and use a plurality of feature information of an article corresponding to the similar image as a plurality of feature information representing the identity of the target object; the plurality of characteristic information includes at least one of a brand, a model, a production date, an original price, and a transaction price.

In one embodiment, the identification module 802 includes: the calculating unit is used for respectively calculating first similarity of the target image and each reference image in the database; a reference article determining unit, configured to determine a reference article corresponding to a number of reference images with a preset value, where the first similarity is highest; the depreciation category determining unit is used for determining depreciation categories corresponding to the preset numerical value reference articles, and taking the depreciation category with the largest number of the corresponding reference articles as the depreciation category corresponding to the target object; and the second determining unit is used for determining characteristic information which represents the identity of the target object and is associated with the depreciation rate based on the depreciation category corresponding to the target object and a preset depreciation coefficient.

In one embodiment, the identification module 802 includes: the matching unit is used for matching the plurality of characteristic information with a preset characteristic list and determining the information grade of each characteristic information; the preset feature list comprises feature information and information levels with mapping relations; and the third determining unit is used for determining at least one piece of feature information in the plurality of pieces of feature information based on the information levels, inputting the feature information into a preset depreciation model, and obtaining feature information which represents the identity of the target object and is associated with the depreciation rate.

In one embodiment, the analysis module 803 includes: the analysis unit is used for carrying out correlation analysis on the characteristic information; the processing unit is used for determining that at least two pieces of feature information are related, eliminating binary features in the at least two pieces of feature information, and fusing at least two pieces of feature information with the related coefficient higher than a second preset threshold value; and the generating unit is used for forming target characteristic information based on the plurality of characteristic information after the elimination and/or fusion processing.

In an embodiment, the analysis module 803 further includes: the mapping unit is used for acquiring a plurality of pieces of feature information after elimination and/or fusion processing, and mapping each piece of feature information to a corresponding preset feature label; and the reporting unit is used for reporting the target characteristic information if at least one preset characteristic label is determined to have no characteristic information corresponding to the preset characteristic label and/or at least one characteristic information corresponding to a target characteristic label in the preset characteristic labels.

In an embodiment, the determining module 804 includes: an execution unit for performing the step of calculating the target object price using a linear regression model trained to converge: determining corresponding linear regression coefficients based on the target characteristic information, and constructing a corresponding linear regression equation; calculating the price of the target object based on the target characteristic information and a linear regression equation; wherein the training of the linear regression model comprises: training a training set input model to obtain a plurality of groups of training coefficients for constructing a linear regression equation; inputting target characteristic information corresponding to an article with a known real price into a linear regression model to obtain a predicted price output by the model; comparing the predicted price with the real price, and if the price error is within a preset range, determining the training coefficient as a linear regression coefficient; the training set comprises target characteristic information corresponding to a plurality of groups of articles with known real prices and real prices thereof.

The image recognition-based item price determining device of the embodiment of the present application may execute an image recognition-based item price determining method provided in the embodiment of the present application, and the implementation principle is similar, the actions performed by each module in the image recognition-based item price determining device of the embodiments of the present application correspond to the steps in the image recognition-based item price determining method of the embodiments of the present application, and for the detailed functional description of each module of the image recognition-based item price determining device, reference may be specifically made to the description in the corresponding image recognition-based item price determining method shown in the foregoing, and details are not repeated here.

Referring now to FIG. 9, shown is a schematic diagram of an electronic device 600 suitable for use in implementing embodiments of the present application. The electronic device in the embodiments of the present application may include, but is not limited to, a device such as a computer. The electronic device shown in fig. 9 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present application.

The electronic device includes: a memory and a processor, wherein the processor may be referred to as the processing device 601 hereinafter, and the memory may include at least one of a Read Only Memory (ROM)602, a Random Access Memory (RAM)603 and a storage device 608 hereinafter, which are specifically shown as follows:

as shown in fig. 9, electronic device 600 may include a processing means (e.g., central processing unit, graphics processor, etc.) 601 that may perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)602 or a program loaded from a storage means 608 into a Random Access Memory (RAM) 603. In the RAM 603, various programs and data necessary for the operation of the electronic apparatus 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.

Generally, the following devices may be connected to the I/O interface 605: input devices 606 including, for example, a touch screen, touch pad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 607 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage 608 including, for example, tape, hard disk, etc.; and a communication device 609. The communication means 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. While fig. 9 illustrates an electronic device 600 having various means, it is to be understood that not all illustrated means are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided.

In particular, according to embodiments of the application, the processes described above with reference to the flow diagrams may be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer program carried on a non-transitory computer readable medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication means 609, or may be installed from the storage means 608, or may be installed from the ROM 602. The computer program, when executed by the processing device 601, performs the above-described functions defined in the methods of the embodiments of the present application.

It should be noted that the computer readable medium mentioned above in the present application may be a computer readable signal medium or a computer readable medium or any combination of the two. A computer readable medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may be any computer readable medium that is not a computer readable medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.

In some embodiments, the clients, servers may communicate using any currently known or future developed network protocol, such as HTTP (HyperText transfer protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communications network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), the Internet (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed network.

The computer readable medium may be embodied in the electronic device; or may exist separately without being assembled into the electronic device.

The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to perform the method steps of the embodiments.

Computer program code for carrying out operations for aspects of the present application may be written in any combination of one or more programming languages, including but not limited to an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The modules or units described in the embodiments of the present application may be implemented by software or hardware. Where the name of a module or unit does not in some cases constitute a limitation of the unit itself, for example, the acquisition module may also be described as a "module that acquires a target image including a target object".

The functions described herein above may be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), systems on a chip (SOCs), Complex Programmable Logic Devices (CPLDs), and the like.

In the context of this application, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The above description is only a preferred embodiment of the application and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the disclosure herein is not limited to the particular combination of features described above, but also encompasses other arrangements formed by any combination of the above features or their equivalents without departing from the spirit of the disclosure. For example, the above features may be replaced with (but not limited to) features having similar functions disclosed in the present application.

Further, while operations are depicted in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are included in the above discussion, these should not be construed as limitations on the scope of the application. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

1. An article price determination method based on image recognition is characterized by comprising the following steps:

acquiring a target image including a target object;

identifying the target image to obtain a plurality of characteristic information representing the identity of the target object;

performing relevance analysis on the plurality of feature information, and determining target feature information in the plurality of feature information based on the analysis result;

and analyzing the target characteristic information through linear regression to determine the price of the target object.

2. The method of claim 1, wherein the identifying the target image and obtaining a plurality of feature information characterizing the identity of the target object comprises:

respectively calculating first similarity of the target image and each reference image in a database;

extracting a first target area image where a preset marker is located in the target image and a second target area image where the preset marker is located in each reference image; calculating a second similarity of the first target area image and the second target area image;

weighting and calculating the first similarity and the second similarity to determine comprehensive similarity;

determining at least one reference image corresponding to the comprehensive similarity higher than a first preset threshold as a similar image, and taking a plurality of feature information of an article corresponding to the similar image as a plurality of feature information representing the identity of the target object; the plurality of characteristic information includes at least one of a brand, a model, a production date, an original price, and a transaction price.

3. The method of claim 1, wherein the identifying the target image and obtaining a plurality of feature information characterizing the identity of the target object comprises:

determining a reference article corresponding to the reference images with the highest first similarity and the preset values;

determining depreciation categories corresponding to the preset numerical value reference articles, and taking the depreciation category with the largest number of the corresponding reference articles as the depreciation category corresponding to the target object;

and determining characteristic information which represents the identity of the target object and is associated with the depreciation rate based on the depreciation category corresponding to the target object and a preset depreciation coefficient.

4. The method of claim 1, wherein the identifying the target image and obtaining a plurality of feature information characterizing the identity of the target object comprises:

matching the plurality of characteristic information with a preset characteristic list, and determining the information grade of each characteristic information; the preset feature list comprises feature information and information levels with mapping relations;

and determining at least one piece of characteristic information in the plurality of pieces of characteristic information based on the information grade, inputting the characteristic information into a preset depreciation model, and obtaining the characteristic information which represents the identity of the target object and is associated with the depreciation rate.

5. The method of claim 1, wherein the performing a correlation analysis on the plurality of feature information, and determining target feature information among the plurality of feature information based on the analysis result comprises:

performing relevance analysis on the plurality of characteristic information;

determining at least two pieces of feature information to be related, excluding binary features in the at least two pieces of feature information, and fusing the at least two pieces of feature information with the related coefficient higher than a second preset threshold;

and forming target characteristic information based on the plurality of characteristic information after the elimination and/or fusion processing.

6. The method according to claim 5, wherein after the forming the target feature information based on the plurality of feature information after the excluding and/or fusing process, the method comprises:

acquiring a plurality of pieces of feature information after elimination and/or fusion processing, and mapping each piece of feature information to a corresponding preset feature tag;

and determining at least one preset feature tag to have no feature information corresponding to the preset feature tag, and/or determining at least one feature information corresponding to a target feature tag in the preset feature tags, and reporting the target feature information.

7. The method of claim 5 or 6, wherein the determining the price of the target object by analyzing the target feature information by linear regression comprises:

performing the step of calculating the target object price using a linear regression model trained to converge: determining corresponding linear regression coefficients based on the target characteristic information, and constructing a corresponding linear regression equation; calculating the price of the target object based on the target characteristic information and a linear regression equation;

8. An article price determining apparatus based on image recognition, comprising:

an acquisition module for acquiring a target image including a target object;

the identification module is used for identifying the target image and acquiring a plurality of characteristic information representing the identity of the target object;

the analysis module is used for carrying out correlation analysis on the characteristic information and determining target characteristic information in the characteristic information based on the analysis result;

and the determining module is used for determining the price of the target object by analyzing the target characteristic information through linear regression.

9. An electronic device, comprising:

the electronic device comprises a memory and a processor;

the memory has stored therein a computer program;

the processor, when executing the computer program, is configured to perform the method of any of claims 1-7.

10. A computer-readable medium, on which a computer program is stored, which, when being executed by a processor, carries out the method of any one of claims 1 to 7.