CN113706513A - Vehicle damage image analysis method, device, equipment and medium based on image detection - Google Patents
Vehicle damage image analysis method, device, equipment and medium based on image detection Download PDFInfo
- Publication number
- CN113706513A CN113706513A CN202111015037.8A CN202111015037A CN113706513A CN 113706513 A CN113706513 A CN 113706513A CN 202111015037 A CN202111015037 A CN 202111015037A CN 113706513 A CN113706513 A CN 113706513A
- Authority
- CN
- China
- Prior art keywords
- image
- vehicle
- vehicle damage
- damage image
- value corresponding
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 36
- 238000003703 image analysis method Methods 0.000 title claims description 8
- 238000004458 analytical method Methods 0.000 claims abstract description 158
- 238000000034 method Methods 0.000 claims abstract description 40
- 230000015654 memory Effects 0.000 claims description 17
- 238000010191 image analysis Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims 1
- 238000013473 artificial intelligence Methods 0.000 abstract description 7
- 206010039203 Road traffic accident Diseases 0.000 description 21
- 238000005516 engineering process Methods 0.000 description 13
- 230000008569 process Effects 0.000 description 9
- 238000012216 screening Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 3
- 230000002349 favourable effect Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000011895 specific detection Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
- G06T2207/30268—Vehicle interior
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the field of artificial intelligence, and discloses an analysis method of a vehicle damage image based on image detection, which comprises the following steps: the method comprises the steps of obtaining a vehicle damage image set to be analyzed and analysis mode information corresponding to the vehicle damage image set, analyzing each vehicle damage image in the vehicle damage image set according to the analysis mode corresponding to the analysis mode information, obtaining a score value corresponding to each vehicle damage image, wherein the analysis mode corresponding to the analysis mode information at least comprises an accessory integrity analysis mode used for determining an accessory integrity value corresponding to each vehicle damage image based on contour pixels and edge pixels in each vehicle damage image, and finally outputting an analysis result corresponding to the vehicle damage image set based on the score value corresponding to each vehicle damage image. Therefore, the method and the device can realize the analysis of the integrity of the accessories in the car damage image based on the contour pixels and the edge pixels in the car damage image, and improve the analysis accuracy of the analysis method of the car damage image.
Description
Technical Field
The invention relates to the field of artificial intelligence, in particular to a method and a device for analyzing a vehicle damage image based on image detection, computer equipment and a storage medium.
Background
Vehicle damage assessment refers to the process of assessing the damage suffered by a vehicle involved in a traffic accident during the traffic accident after the traffic accident occurs, so as to estimate the cost required for repairing the damage. In the actual vehicle damage assessment process, it is usually necessary for a damage assessment person to take hundreds of pictures of the accident site (e.g., the traffic environment of the accident site, the damaged part of the vehicle involved in the accident, the overall vehicle condition of the vehicle involved in the accident, etc.) at the scene of the traffic accident, then screen out the pictures which have reference significance for vehicle damage assessment from the hundreds of pictures taken, and finally assess the cost required for repairing the damage by observing the pictures.
At present, when images with reference significance for vehicle damage assessment are screened out, automatic analysis of vehicle damage images by using a computer technology can be achieved to assist users in screening. However, when a computer technology is used to automatically analyze the car damage image, how to analyze the car damage image by using a computer to ensure the accuracy of the analysis result is a technical problem that needs to be solved. Therefore, the analysis accuracy of the current analysis method for the vehicle damage image still has a space for further improvement.
Disclosure of Invention
The invention aims to solve the technical problem that the analysis accuracy of the conventional vehicle damage image analysis method is low.
In order to solve the technical problem, a first aspect of the present invention discloses a method for analyzing a vehicle damage image based on image detection, where the method includes:
acquiring a vehicle loss image set to be analyzed, wherein the vehicle loss image set comprises a plurality of vehicle loss images;
acquiring analysis mode information corresponding to the vehicle damage image set;
analyzing each vehicle loss image in the vehicle loss image set according to a first analysis mode corresponding to the analysis mode information to obtain contour pixels and edge pixels in the vehicle loss image;
determining an accessory integrity value corresponding to each car damage image based on the contour pixels and the edge pixels in each car damage image;
determining a score value corresponding to each vehicle damage image based on the accessory integrity value corresponding to each vehicle damage image;
and outputting vehicle damage image analysis results corresponding to the vehicle damage image set based on the score value corresponding to each vehicle damage image so as to screen out the target vehicle damage image.
The invention discloses a vehicle damage image analysis device based on image detection in a second aspect, which comprises:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring a vehicle loss image set to be analyzed, and the vehicle loss image set comprises a plurality of vehicle loss images;
the acquisition module is further used for acquiring analysis mode information corresponding to the vehicle damage image set;
the analysis module is used for analyzing each vehicle loss image in the vehicle loss image set according to a first analysis mode corresponding to the analysis mode information to obtain contour pixels and edge pixels in the vehicle loss image;
the determining module is used for determining an accessory integrity value corresponding to each car damage image based on the contour pixels and the edge pixels in each car damage image;
the determining module is further configured to determine a score value corresponding to each vehicle damage image based on the accessory integrity value corresponding to each vehicle damage image;
and the output module is used for outputting the vehicle damage image analysis result corresponding to the vehicle damage image set based on each rating value corresponding to the vehicle damage image so as to screen out the target vehicle damage image. A third aspect of the present invention discloses a computer apparatus, comprising:
a memory storing executable program code;
a processor coupled to the memory;
the processor calls the executable program code stored in the memory to execute part or all of the steps in the analysis method of the vehicle damage image based on the image detection disclosed by the first aspect of the invention.
In a fourth aspect of the present invention, a computer storage medium is disclosed, which stores computer instructions for performing some or all of the steps of the method for analyzing a vehicle damage image based on image detection disclosed in the first aspect of the present invention when the computer instructions are called.
In the embodiment of the invention, firstly, analysis mode information corresponding to a vehicle damage image set to be analyzed and the vehicle damage image set is obtained, then each vehicle damage image in the vehicle damage image set is analyzed according to the analysis mode corresponding to the analysis mode information to obtain a score value corresponding to each vehicle damage image, wherein the analysis mode corresponding to the analysis mode information at least comprises an accessory integrity analysis mode for determining an accessory integrity value corresponding to each vehicle damage image based on a contour pixel and an edge pixel in each vehicle damage image, and finally, an analysis result corresponding to the vehicle damage image set is output based on the score value corresponding to each vehicle damage image to assist a user in screening a target vehicle damage image, so that the integrity of accessories in the vehicle damage image can be analyzed based on the contour pixel and the edge pixel in the vehicle damage image, and the vehicle damage image can be analyzed from the dimension of the integrity of the accessories in the vehicle damage image, therefore, the analysis result of the vehicle damage image can be more accurate.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic flowchart of an analysis method of a vehicle damage image based on image detection according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an analysis apparatus for a vehicle damage image based on image detection according to an embodiment of the present invention;
FIG. 3 is a schematic structural diagram of a computer device according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a computer storage medium according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The terms "first," "second," and the like in the description and claims of the present invention and in the above-described drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "include" and "have," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, article, or article that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or inherent to such process, method, article, or article.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the invention. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.
The embodiment of the application can acquire and process related data based on an artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result.
The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like.
The invention discloses a vehicle damage image analysis method based on image detection, a device, a computer device and a storage medium, firstly acquiring analysis mode information corresponding to a vehicle damage image set to be analyzed and the vehicle damage image set, then analyzing each vehicle damage image in the vehicle damage image set according to the analysis mode corresponding to the analysis mode information to obtain a score value corresponding to each vehicle damage image, wherein the analysis mode corresponding to the analysis mode information at least comprises an accessory integrity analysis mode for determining an accessory integrity value corresponding to each vehicle damage image based on contour pixels and edge pixels in each vehicle damage image, and finally outputting an analysis result corresponding to the vehicle damage image set based on the score value corresponding to each vehicle damage image to assist a user in screening a target vehicle damage image, so that the analysis of the integrity of accessories in the vehicle damage image can be realized based on the contour pixels and the edge pixels in the vehicle damage image, the car damage image is analyzed from the dimension of the integrity of the accessories in the car damage image, so that the analysis result of the car damage image can be more accurate. The following are detailed below.
Example one
Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a method for analyzing a vehicle damage image based on image detection according to an embodiment of the present invention. As shown in fig. 1, the analysis method of the vehicle damage image based on image detection may include the following operations:
101. the method comprises the steps of obtaining a vehicle loss image set to be analyzed, wherein the vehicle loss image set comprises a plurality of vehicle loss images.
In the step 101, the set of damage images to be analyzed may be pre-stored in the vehicle insurance system. Specifically, the loss assessment personnel can upload images captured in the traffic accident as vehicle loss images to the vehicle insurance system. Usually, in a traffic accident, the loss assessment personnel will arrive at the scene of the traffic accident and take hundreds of images, and then upload the images to the vehicle insurance system as a vehicle loss image set of the traffic accident. Subsequently, when the loss assessment personnel need to assess the loss of the traffic accident, the loss image set of the traffic accident can be directly extracted from the vehicle insurance system so as to be used for vehicle loss assessment.
102. And acquiring analysis mode information corresponding to the vehicle damage image set.
In step 102, the analysis method information corresponding to the vehicle damage image set is used to record the analysis method to be used for analyzing the vehicle damage image set. Specifically, the analysis mode information corresponding to the vehicle damage image set may be determined in the following two modes:
(1) and setting analysis mode information corresponding to the vehicle damage image set by the damage assessment personnel when the vehicle damage image set is uploaded to the vehicle insurance system. For example, in a certain traffic accident, the image definition of the vehicle damage image of the traffic accident needs to be analyzed subsequently, and when the damage assessment personnel uploads the vehicle damage image set to the vehicle insurance system, the analysis mode information corresponding to the traffic accident can be set as the analysis needing to be performed on the image definition.
(2) When a loss assessment person uploads a vehicle loss image set to a vehicle insurance system, a traffic accident type corresponding to the vehicle loss image set is set, and then analysis mode information corresponding to the traffic accident type is inquired in a preset analysis mode information table to serve as analysis mode information corresponding to the vehicle loss image set. The analysis mode information table may store analysis mode information corresponding to different traffic accident types in advance.
103. And analyzing each vehicle loss image in the vehicle loss image set according to a first analysis mode corresponding to the analysis mode information to obtain contour pixels and edge pixels in the vehicle loss image.
In step 103, the contour refers to a connected and closed curve in the image, and the edge refers to an unconnected and open curve in the image. The outline curve and the edge curve in the car damage image can be extracted through an opencv tool, pixels in the extracted outline curve are outline pixels, and pixels in the extracted edge curve are edge pixels. The analysis mode of the vehicle damage image can comprise a first analysis mode, a second analysis mode and a third analysis mode, and the first analysis mode, the second analysis mode and the third analysis mode are respectively used for analyzing the accessory integrity value, the image size matching value and the image definition value of the image. The user can set the analysis performed on the car damage image by setting the analysis mode information.
At present, when an artificial intelligence technology is used for analyzing a vehicle loss image, the analysis mode is generally fixed and single, for example, only the definition of the vehicle loss image is analyzed, only the size of the vehicle loss image is analyzed, and the like. However, in actual damage assessment, the analysis requirements of the vehicle damage image set are usually complicated and varied. For example, the analysis requirements for the loss assessment photos for different types of traffic accidents are usually different, for example, the analysis for the definition of the loss assessment photos is not usually required for the traffic accidents with determined license plate numbers of responsible subjects, only the integrity of accessories in the loss assessment photos needs to be analyzed, and the analysis for the definition of the loss assessment photos needs to be performed for the traffic accidents with no determined license plate numbers of responsible subjects (because the license plate numbers of responsible subjects need to be extracted from the loss assessment photos, the loss assessment photos need to be guaranteed to have enough definition to recognize the license plate numbers). In this case, it is obviously not appropriate to use a fixed single analysis method for both of the sets of the loss images of the two types of traffic accidents. Therefore, the existing analysis method of the vehicle damage image also has the problems that the analysis mode is fixed and single, and the method cannot be flexibly adapted to the actual application scene. In the embodiment of the invention, different analysis modes are set for different vehicle loss image sets, and then different analysis modes are adopted for different vehicle loss image sets to be analyzed (a specific analysis process is described later), so that the vehicle loss image analysis method with rich analysis modes can be provided, and the method can be more flexibly suitable for various practical application scenes. In addition, in the analysis method of the car damage image according to the embodiment of the invention, an analysis mode for analyzing the integrity of the accessories in the car damage image based on the contour pixels and the edge pixels in the car damage image is also provided, and the score value of the car damage image is determined based on the accessory integrity value of the car damage image. In the prior art, it is an urgent technical problem to analyze a vehicle damage image from what dimension to ensure the accuracy of an analysis result. In the embodiment of the invention, the integrity of the accessories in the vehicle damage image is analyzed based on the contour pixels and the edge pixels in the vehicle damage image, and the vehicle damage image is analyzed from the dimension of the integrity of the accessories in the vehicle damage image, so that the analysis result of the vehicle damage image is more accurate. The process of analyzing the integrity of the parts in the specific damage image will be described later.
104. And determining an accessory integrity value corresponding to each car damage image based on the contour pixels and the edge pixels in each car damage image.
In the step 104, when the vehicle damage image is analyzed in the first analysis manner, the contour pixels and the edge pixels extracted from the vehicle damage image may be used to calculate the accessory integrity value corresponding to the vehicle damage image, and a specific calculation process will be described later.
105. And determining the score value corresponding to the vehicle damage image based on the accessory integrity value corresponding to each vehicle damage image.
In step 105, after the component integrity value is calculated, the component integrity value may be used as a factor to calculate the corresponding score value of the car damage image.
106. And outputting vehicle damage image analysis results corresponding to the vehicle damage image set based on the score value corresponding to each vehicle damage image so as to screen out the target vehicle damage image.
In step 106, after the analysis of the vehicle damage image set is completed, a score value corresponding to each vehicle damage image in the vehicle damage image set can be obtained. At this time, each car damage image and the score value corresponding to the car damage image may be presented to the user together in the interactive interface, and the score value corresponding to the car damage image may be marked on the car damage image on the interactive interface (i.e., an analysis result corresponding to the car damage image set is output). Therefore, the user can visually observe the score of each vehicle damage image, and then can select a proper target vehicle damage image by referring to the score of the vehicle damage image so as to perform subsequent vehicle damage assessment work. Optionally, all the vehicle damage images may also be sorted in the order of the score values from high to low, and presented to the user on the interactive interface according to the sorting (i.e., the analysis result corresponding to the vehicle damage image set is output). Therefore, the user can quickly and accurately select a proper target vehicle damage image according to the ranking of the vehicle damage images so as to perform subsequent vehicle damage assessment work.
It can be seen that, when the analysis method of the car damage image based on image detection described in fig. 1 is implemented, firstly, the car damage image set to be analyzed and the analysis mode information corresponding to the car damage image set are obtained, then analyzing each vehicle damage image in the vehicle damage image set according to the analysis mode corresponding to the analysis mode information to obtain a score value corresponding to each vehicle damage image, finally outputting an analysis result corresponding to the vehicle damage image set based on the score value corresponding to each vehicle damage image to assist a user in screening a target vehicle damage image, therefore, when the user screens out the photos with reference significance for vehicle damage assessment from the vehicle damage image set, the automatic analysis of the vehicle damage image is realized by utilizing the computer technology, the method is used for assisting the user in screening the vehicle loss images, and improving the analysis efficiency and the analysis quality of the analysis method of the vehicle loss images during vehicle loss image screening. In addition, different analysis modes are set for different vehicle loss image sets, and then different analysis modes are adopted for different vehicle loss image sets to analyze, so that the vehicle loss image analysis method with rich analysis modes can be provided, various practical application scenes can be flexibly adapted, and the automatic analysis effect of the vehicle loss image is improved. The integrity of the accessories in the car damage image is analyzed based on the contour pixels and the edge pixels in the car damage image, and the car damage image is analyzed from the dimension of the integrity of the accessories in the car damage image, so that the analysis result of the car damage image can be more accurate.
In an optional embodiment, the determining, based on the contour pixel and the edge pixel in each of the car damage images, an accessory integrity value corresponding to the car damage image includes:
detecting whether a target contour pixel corresponding to the vehicle damage image exists in contour pixels of each vehicle damage image in the vehicle damage image set;
when detecting that the target contour pixels corresponding to the vehicle damage image exist in the contour pixels of the vehicle damage image, setting a preset target accessory integrity value as an accessory integrity value corresponding to the vehicle damage image;
when the target contour pixel does not exist in the contour pixels of the vehicle damage image, calculating the ratio of the number of the edge pixels in the vehicle damage image to the number of the preset complete pixels corresponding to the vehicle damage image to be used as the accessory integrity value corresponding to the vehicle damage image.
In this alternative embodiment, the target contour pixel corresponding to each car damage image may be preset, for example, when one car damage image is an image for recording the damage condition of the wheels of the automobile, the target contour pixel corresponding to the car damage image may be set as a pixel in a circular ring shape. When it is detected that a pixel (i.e., a target contour pixel) in the same circular ring shape exists in the car damage image (a specific detection process, which is described in detail later), it is indicated that an automobile part with a complete wheel exists in the car damage image, that is, the part integrity value corresponding to the car damage image may be set to a full value (i.e., a target part integrity value), for example, the part integrity value corresponding to the car damage image is set to 1. When it is detected that there is no pixel (i.e., a target contour pixel) in the same annular shape in the vehicle loss image, it is indicated that there is no automobile accessory with a complete wheel in the vehicle loss image, at this time, an edge curve in the vehicle loss image may be extracted, and then a ratio of the number of pixels in the edge curve (i.e., the number of pixels of the edge pixel) to the number of complete pixels corresponding to the preset vehicle loss image is calculated to be used as an accessory integrity value corresponding to the vehicle loss image. For example, if there is only one third wheel in the damage image, the final calculated integrity value of the accessory corresponding to the damage image may be one third. Alternatively, the accessory integrity value corresponding to the car damage image can be directly used as the score value corresponding to the car damage image.
It can be seen that, by implementing the optional embodiment, when a target contour pixel corresponding to the vehicle damage image exists in the vehicle damage image, the preset target accessory integrity value is set as the accessory integrity value corresponding to the vehicle damage image, and when the target contour pixel corresponding to the vehicle damage image does not exist in the vehicle damage image, a ratio of the number of pixels of the edge pixel in the vehicle damage image to the preset number of complete pixels corresponding to the vehicle damage image is calculated to serve as the accessory integrity value corresponding to the vehicle damage image, so that the accessory integrity value corresponding to the vehicle damage image can be calculated by detecting the contour pixel or the edge pixel in the vehicle damage image, and then the accessory integrity value corresponding to the vehicle damage image is calculated, thereby realizing the analysis of the integrity of the vehicle accessories in the vehicle damage image.
In an optional embodiment, the detecting whether a target contour pixel corresponding to each loss image exists in the contour pixels of each loss image in the loss image set includes:
calculating the ratio of the pixel number of the contour pixels of the vehicle damage image to the preset pixel number of the target contour pixels corresponding to the vehicle damage image;
judging whether the ratio is within a preset ratio range or not;
when the ratio is judged to be within the ratio range, determining that target contour pixels corresponding to the vehicle loss image exist in the vehicle loss image;
and when the ratio is judged not to be in the ratio range, determining that the target contour pixels corresponding to the vehicle loss image do not exist in the vehicle loss image.
In this alternative embodiment, the outlines of the different types of accessories in the damage image are generally different, for example, if the wheel is an annular outline, the rearview mirror is a rectangle-like outline, and the outlines of the accessories are different, the number of pixels occupied by the outlines of a complete accessory in the damage image is also different. Therefore, different pixel numbers of the target contour pixels can be preset for the vehicle loss images of different types of accessories, and then whether the target contour pixels corresponding to the vehicle loss images exist in the vehicle loss images or not (that is, whether complete accessories exist in the vehicle loss images or not) is judged by judging whether the pixel numbers of the contour pixels of the vehicle loss images are matched with the pixel numbers of the target contour pixels corresponding to the preset vehicle loss images or not. For example, for a vehicle loss image of a wheel, the number of pixels of a target contour pixel corresponding to the vehicle loss image may be preset to 1000, and the ratio range may be set to 0.9-1.1, and if there is a wheel in the vehicle loss image, the number of pixels of the extracted contour pixel may also be approximately distributed around 1000, for example, the number of pixels of the extracted contour pixel is 960, at this time, the ratio of the number of pixels of the extracted contour pixel to the number of pixels of the target contour pixel corresponding to the vehicle loss image is calculated to be 0.96, and is within the ratio range, so it may be determined that there is a target contour pixel corresponding to the vehicle loss image in the vehicle loss image (i.e., there is a complete wheel). If only one third of wheels exist in the vehicle damage image, the number of the extracted contour pixels may be 300, and at this time, the calculated ratio is 0.3 and is not within the ratio range, so that it can be determined that the target contour pixels corresponding to the vehicle damage image do not exist in the vehicle damage image (i.e., a complete wheel does not exist).
Therefore, the optional embodiment is implemented, the contour pixels of the vehicle loss image are extracted, the ratio of the number of the contour pixels of the vehicle loss image to the number of the target contour pixels corresponding to the vehicle loss image is calculated, and finally whether the target contour pixels exist in the vehicle loss image is determined according to whether the ratio is in the ratio range, so that the target contour pixels in the vehicle loss image are detected, and a basis is provided for the analysis of the integrity of the vehicle parts in the vehicle loss image.
In an optional embodiment, before determining the score value corresponding to each of the car damage images based on the accessory integrity value corresponding to the car damage image, the method further includes:
according to a second analysis mode corresponding to the analysis mode information, calculating an image size matching value of each vehicle loss image in the vehicle loss image set through the following formula:
wherein, a is an image size matching value of the vehicle damage image, b is a width pixel value of the vehicle damage image, c is a preset standard width pixel value, d is a length pixel value of the vehicle damage image, and e is a preset standard length pixel value;
and determining the score value corresponding to the vehicle damage image based on the accessory integrity value corresponding to each vehicle damage image, wherein the score value comprises the following steps:
and determining the score value corresponding to the vehicle damage image based on the accessory integrity value and the image size matching value corresponding to each vehicle damage image.
In this optional embodiment, when the vehicle damage image is analyzed, whether the image size of the vehicle damage image meets a preset image size may also be analyzed, that is, the second analysis mode. For example, if the width pixel value of the car damage image is 90, the length pixel value is 110, and the preset standard width pixel value and the preset standard length pixel value are both 100, the calculated image size matching value is:
the image size matching value calculated through the formula can represent the matching degree of the image size of the vehicle damage image and the preset image size. Optionally, a weighted sum of the accessory integrity value and the image size matching value corresponding to the car damage image may be used as the score value corresponding to the car damage image.
Therefore, when the optional embodiment is implemented, when the vehicle loss image is analyzed, an image size matching value of the vehicle loss image, which can represent the matching degree between the image size of the vehicle loss image and the preset image size, is calculated according to the preset formula, so that the analysis of the image size of the vehicle loss image can be realized, the analysis dimensionality of the analysis method of the vehicle loss image is increased, and the analysis effect is improved.
In an optional embodiment, before determining the score value corresponding to each of the car damage images based on the accessory integrity value and the image size matching value corresponding to the car damage image, the method further includes:
analyzing each vehicle loss image in the vehicle loss image set based on a preset image definition analysis model according to a third analysis mode corresponding to the analysis mode information to obtain an image definition value corresponding to the vehicle loss image;
and determining the score value corresponding to the vehicle damage image based on the accessory integrity value and the image size matching value corresponding to each vehicle damage image, wherein the score value comprises the following steps:
and determining the score value corresponding to the vehicle damage image based on the accessory integrity value, the image size matching value and the image definition value corresponding to each vehicle damage image.
In this alternative embodiment, the image sharpness analysis model may be a RankIQA model. Different levels and types of ordered distorted images can be generated by performing image processing transformation on the determined definition picture, so that a large data set can be obtained, and a wider and deeper network can be selected for training. Specifically, the Siamese network can be selected to learn the representation characteristics for generating the data ordering relationship, and then the knowledge represented in the trained Siamese network is transferred to the traditional CNN, so that the image definition value of a single image can be predicted.
Therefore, when the optional embodiment is implemented, when the vehicle loss image is analyzed, the image definition value corresponding to the vehicle loss image is analyzed through the preset image definition analysis model, so that the analysis of the image definition of the vehicle loss image can be realized, the analysis dimensionality of the analysis method of the vehicle loss image is further increased, and the analysis effect is improved.
In an optional embodiment, the determining a score value corresponding to each of the car damage images based on the accessory integrity value, the image size matching value, and the image clarity value corresponding to the car damage image includes:
calculating a weighted sum of an accessory integrity value, an image size matching value and an image definition value corresponding to the vehicle damage image based on a preset weight value to serve as an original score value corresponding to the vehicle damage image;
calculating the score value corresponding to the car damage image based on the original score value corresponding to the car damage image by the following formula:
wherein f is the score value corresponding to the vehicle damage image, g is the original score value corresponding to the vehicle damage image, max is the original score value with the largest value among the original score values corresponding to all the vehicle damage images, and min is the original score value with the smallest value among the original score values corresponding to all the vehicle damage images.
In this alternative embodiment, when the weighted sum of the component integrity value, the image size matching value, and the image clarity value corresponding to the damage image is used as the original score value corresponding to the damage image, if the original score value is directly used as the score value of the damage image, since the distribution of the values of the original score value is usually quite random (for example, one original score value may be 10, and the other original score value may be 1000), it is not favorable for the user to observe the analysis result, and therefore, it is not favorable for the user to screen the damage image according to the analysis result. Through the formula, the original score value can be mapped to the [0,1] interval to be used as the score value, and meanwhile, the score value obtained through mapping also keeps the function of representing the score of the vehicle damage image by the original score value, so that the finally obtained score value is more convenient to be directly observed by a user, and the vehicle damage image can be screened by the user according to an analysis result.
Therefore, by implementing the optional embodiment, after the weighting sum of the accessory integrity value, the image size matching value and the image definition value corresponding to the car damage image is used as the original score value corresponding to the car damage image, the original score value is mapped to the specified numerical value interval through the preset formula to be used as the score value, so that the finally obtained score value can be more conveniently and directly observed by a user, and the car damage image can be screened by the user according to the analysis result.
In an optional embodiment, before the acquiring the set of vehicle damage images to be analyzed, the method further includes:
acquiring an original vehicle damage image uploaded by a user;
analyzing the original vehicle damage image according to all analysis modes corresponding to the analysis mode information to obtain a score value corresponding to the original vehicle damage image;
judging whether the score value corresponding to the original vehicle damage image is larger than a preset score value threshold value or not;
when the score value corresponding to the original vehicle damage image is judged to be larger than the score value threshold value, adding the original vehicle damage image to a vehicle damage image set;
and when judging that the score value corresponding to the original vehicle damage image is not greater than the score value threshold value, outputting a warning prompt to the user, wherein the warning prompt is used for prompting the user to upload the original vehicle damage image again.
In this alternative embodiment, the damage images in the set of damage images may be uploaded by the damage rater. Specifically, the loss assessment personnel can use the mobile terminal to shoot the vehicle loss image (i.e. the original vehicle loss image) at the scene of the traffic accident and then upload the vehicle loss image to the vehicle insurance system. After the user uploads the original vehicle damage image, the original vehicle damage image can be analyzed once to obtain the score value of the original vehicle damage image. If the score value of the original vehicle damage image is higher than a preset score value threshold value, the original vehicle damage image is added into the vehicle damage image set, if the score value of the original vehicle damage image is not higher than the preset score value threshold value, the image quality of the original vehicle damage image is indicated to be unqualified, and a warning prompt is output to a user to prompt the user to upload the original vehicle damage image again. Therefore, the vehicle damage images uploaded to the vehicle damage image set can be screened according to the score values, and the image quality of the vehicle damage image set is improved.
It can be seen that, by implementing the optional embodiment, after the user uploads the original vehicle loss image, the uploaded original vehicle loss image is analyzed to obtain the score of the original vehicle loss image, if the score of the original vehicle loss image is greater than the score threshold value, the original vehicle loss image is added to the vehicle loss image set, and if the score of the original vehicle loss image is not greater than the score threshold value, the user is prompted to upload the original vehicle loss image again, so that the vehicle loss image uploaded to the vehicle loss image set can be screened according to the score value, and the image quality of the vehicle loss image set is favorably improved.
Optionally, it is also possible: uploading analysis information of the vehicle loss image based on image detection analysis method to a block chain.
Specifically, the analysis information of the vehicle damage image based on image detection is obtained by running the analysis method of the vehicle damage image based on image detection, and is used for recording the analysis condition of the vehicle damage image based on image detection, such as the acquired vehicle damage image set, the analysis mode information, the score value of the vehicle damage image, and the like. The analysis information of the vehicle damage image based on the image detection is uploaded to the block chain, so that the safety and the fair transparency to users can be guaranteed. The user may download the analysis information of the image-based vehicle damage image from the blockchain, so as to verify whether the analysis information of the image-based vehicle damage image analysis method is tampered. The blockchain referred to in this example is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, encryption algorithm, and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
Example two
Referring to fig. 2, fig. 2 is a schematic structural diagram of an analysis apparatus for a vehicle damage image based on image detection according to an embodiment of the present invention. As shown in fig. 2, the apparatus for analyzing a vehicle damage image based on image detection may include:
an obtaining module 201, configured to obtain a vehicle loss image set to be analyzed, where the vehicle loss image set includes a plurality of vehicle loss images;
the obtaining module 201 is further configured to obtain analysis mode information corresponding to the vehicle damage image set;
the analysis module 202 is configured to analyze each vehicle loss image in the vehicle loss image set according to a first analysis mode corresponding to the analysis mode information to obtain a contour pixel and an edge pixel in the vehicle loss image;
the determining module 203 is configured to determine an accessory integrity value corresponding to each of the car damage images based on the contour pixels and the edge pixels in the car damage image;
the determining module 203 is further configured to determine a score value corresponding to each vehicle damage image based on the accessory integrity value corresponding to each vehicle damage image;
and the output module 204 is configured to output a vehicle damage image analysis result corresponding to the vehicle damage image set based on the score value corresponding to each vehicle damage image so as to screen out a target vehicle damage image.
For the specific description of the analysis apparatus for the vehicle damage image based on the image detection, reference may be made to the specific description of the analysis method for the vehicle damage image based on the image detection, and for avoiding repetition, details are not repeated here.
EXAMPLE III
Referring to fig. 3, fig. 3 is a schematic structural diagram of a computer device according to an embodiment of the present invention. As shown in fig. 3, the computer apparatus may include:
a memory 301 storing executable program code;
a processor 302 connected to the memory 301;
the processor 302 calls the executable program code stored in the memory 301 to execute the steps in the analysis method of the car damage image based on image detection disclosed in the embodiment of the present invention.
Example four
Referring to fig. 4, an embodiment of the present invention discloses a computer storage medium 401, where the computer storage medium 401 stores computer instructions, and the computer instructions, when called, are used to execute steps in an analysis method for a car damage image based on image detection disclosed in an embodiment of the present invention.
The above-described embodiments of the apparatus are merely illustrative, and the modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, may be located in one place, or may be distributed on a plurality of network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above detailed description of the embodiments, those skilled in the art will clearly understand that the embodiments may be implemented by software plus a necessary general hardware platform, and may also be implemented by hardware. Based on such understanding, the above technical solutions may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, where the storage medium includes a Read-Only Memory (ROM), a Random Access Memory (RAM), a Programmable Read-Only Memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), a One-time Programmable Read-Only Memory (OTPROM), an Electrically Erasable Programmable Read-Only Memory (EEPROM), a Compact Disc-Read-Only Memory (CD-ROM), or other disk memories, CD-ROMs, or other magnetic disks, A tape memory, or any other medium readable by a computer that can be used to carry or store data.
Finally, it should be noted that: the method, the apparatus, the computer device and the storage medium for analyzing the car damage image based on image detection disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, and are only used for illustrating the technical solution of the present invention, not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art; the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A vehicle damage image analysis method based on image detection is characterized by comprising the following steps:
acquiring a vehicle loss image set to be analyzed, wherein the vehicle loss image set comprises a plurality of vehicle loss images;
acquiring analysis mode information corresponding to the vehicle damage image set;
analyzing each vehicle loss image in the vehicle loss image set according to a first analysis mode corresponding to the analysis mode information to obtain contour pixels and edge pixels in the vehicle loss image;
determining an accessory integrity value corresponding to each car damage image based on the contour pixels and the edge pixels in each car damage image;
determining a score value corresponding to each vehicle damage image based on the accessory integrity value corresponding to each vehicle damage image;
and outputting vehicle damage image analysis results corresponding to the vehicle damage image set based on the score value corresponding to each vehicle damage image so as to screen out the target vehicle damage image.
2. The method for analyzing a car damage image based on image detection according to claim 1, wherein the determining a fitting integrity value corresponding to each car damage image based on the contour pixels and the edge pixels in the car damage image comprises:
detecting whether a target contour pixel corresponding to the vehicle damage image exists in contour pixels of each vehicle damage image in the vehicle damage image set;
when detecting that the target contour pixels corresponding to the vehicle damage image exist in the contour pixels of the vehicle damage image, setting a preset target accessory integrity value as an accessory integrity value corresponding to the vehicle damage image;
when the target contour pixel does not exist in the contour pixels of the vehicle damage image, calculating the ratio of the number of the edge pixels in the vehicle damage image to the number of the preset complete pixels corresponding to the vehicle damage image to be used as the accessory integrity value corresponding to the vehicle damage image.
3. The method for analyzing a vehicle damage image based on image detection according to claim 2, wherein the detecting whether a target contour pixel corresponding to each vehicle damage image exists in the contour pixels of each vehicle damage image in the vehicle damage image set comprises:
calculating the ratio of the pixel number of the contour pixels of the vehicle damage image to the preset pixel number of the target contour pixels corresponding to the vehicle damage image;
judging whether the ratio is within a preset ratio range or not;
when the ratio is judged to be within the ratio range, determining that target contour pixels corresponding to the vehicle loss image exist in the vehicle loss image;
and when the ratio is judged not to be in the ratio range, determining that the target contour pixels corresponding to the vehicle loss image do not exist in the vehicle loss image.
4. The method for analyzing vehicle damage images based on image detection according to any one of claims 1-3, wherein before determining the score value corresponding to each vehicle damage image based on the accessory integrity value corresponding to the vehicle damage image, the method further comprises:
according to a second analysis mode corresponding to the analysis mode information, calculating an image size matching value of each vehicle loss image in the vehicle loss image set through the following formula:
wherein, a is an image size matching value of the vehicle damage image, b is a width pixel value of the vehicle damage image, c is a preset standard width pixel value, d is a length pixel value of the vehicle damage image, and e is a preset standard length pixel value;
and determining the score value corresponding to the vehicle damage image based on the accessory integrity value corresponding to each vehicle damage image, wherein the score value comprises the following steps:
and determining the score value corresponding to the vehicle damage image based on the accessory integrity value and the image size matching value corresponding to each vehicle damage image.
5. The method for analyzing vehicle damage images based on image detection as claimed in claim 4, wherein before determining the score value corresponding to each vehicle damage image based on the accessory integrity value and the image size matching value corresponding to the vehicle damage image, the method further comprises:
analyzing each vehicle loss image in the vehicle loss image set based on a preset image definition analysis model according to a third analysis mode corresponding to the analysis mode information to obtain an image definition value corresponding to the vehicle loss image;
and determining the score value corresponding to the vehicle damage image based on the accessory integrity value and the image size matching value corresponding to each vehicle damage image, wherein the score value comprises the following steps:
and determining the score value corresponding to the vehicle damage image based on the accessory integrity value, the image size matching value and the image definition value corresponding to each vehicle damage image.
6. The method for analyzing vehicle damage images based on image detection according to claim 5, wherein the determining the score value corresponding to each vehicle damage image based on the accessory integrity value, the image size matching value and the image clarity value corresponding to the vehicle damage image comprises:
calculating a weighted sum of an accessory integrity value, an image size matching value and an image definition value corresponding to the vehicle damage image based on a preset weight value to serve as an original score value corresponding to the vehicle damage image;
calculating the score value corresponding to the car damage image based on the original score value corresponding to the car damage image by the following formula:
wherein f is the score value corresponding to the vehicle damage image, g is the original score value corresponding to the vehicle damage image, max is the original score value with the largest value among the original score values corresponding to all the vehicle damage images, and min is the original score value with the smallest value among the original score values corresponding to all the vehicle damage images.
7. The method for analyzing vehicle damage images based on image detection according to claim 1, wherein before the acquiring the set of vehicle damage images to be analyzed, the method further comprises:
acquiring an original vehicle damage image uploaded by a user;
analyzing the original vehicle damage image according to all analysis modes corresponding to the analysis mode information to obtain a score value corresponding to the original vehicle damage image;
judging whether the score value corresponding to the original vehicle damage image is larger than a preset score value threshold value or not;
when the score value corresponding to the original vehicle damage image is judged to be larger than the score value threshold value, adding the original vehicle damage image to a vehicle damage image set;
and when judging that the score value corresponding to the original vehicle damage image is not greater than the score value threshold value, outputting a warning prompt to the user, wherein the warning prompt is used for prompting the user to upload the original vehicle damage image again.
8. An apparatus for analyzing a vehicle damage image based on image detection, the apparatus comprising:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring a vehicle loss image set to be analyzed, and the vehicle loss image set comprises a plurality of vehicle loss images;
the acquisition module is further used for acquiring analysis mode information corresponding to the vehicle damage image set;
the analysis module is used for analyzing each vehicle loss image in the vehicle loss image set according to a first analysis mode corresponding to the analysis mode information to obtain contour pixels and edge pixels in the vehicle loss image;
the determining module is used for determining an accessory integrity value corresponding to each car damage image based on the contour pixels and the edge pixels in each car damage image;
the determining module is further configured to determine a score value corresponding to each vehicle damage image based on the accessory integrity value corresponding to each vehicle damage image;
and the output module is used for outputting the vehicle damage image analysis result corresponding to the vehicle damage image set based on each rating value corresponding to the vehicle damage image so as to screen out the target vehicle damage image.
9. A computer device, characterized in that the computer device comprises:
a memory storing executable program code;
a processor coupled to the memory;
the processor calls the executable program code stored in the memory to execute the method for analyzing the vehicle damage image based on image detection according to any one of claims 1 to 7.
10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out a method of image detection-based analysis of a vehicle damage image according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111015037.8A CN113706513A (en) | 2021-08-31 | 2021-08-31 | Vehicle damage image analysis method, device, equipment and medium based on image detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111015037.8A CN113706513A (en) | 2021-08-31 | 2021-08-31 | Vehicle damage image analysis method, device, equipment and medium based on image detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113706513A true CN113706513A (en) | 2021-11-26 |
Family
ID=78658231
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111015037.8A Pending CN113706513A (en) | 2021-08-31 | 2021-08-31 | Vehicle damage image analysis method, device, equipment and medium based on image detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113706513A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114462553A (en) * | 2022-04-12 | 2022-05-10 | 之江实验室 | Image labeling and element extraction method and system for car insurance fraud prevention |
| CN115810134A (en) * | 2023-02-14 | 2023-03-17 | 之江实验室 | Image acquisition quality inspection method, system and device for preventing car insurance from cheating |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110942072A (en) * | 2019-12-31 | 2020-03-31 | 北京迈格威科技有限公司 | Quality evaluation-based quality scoring and detecting model training and detecting method and device |
| CN112381840A (en) * | 2020-11-27 | 2021-02-19 | 深源恒际科技有限公司 | Method and system for marking vehicle appearance parts in loss assessment video |
-
2021
- 2021-08-31 CN CN202111015037.8A patent/CN113706513A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110942072A (en) * | 2019-12-31 | 2020-03-31 | 北京迈格威科技有限公司 | Quality evaluation-based quality scoring and detecting model training and detecting method and device |
| CN112381840A (en) * | 2020-11-27 | 2021-02-19 | 深源恒际科技有限公司 | Method and system for marking vehicle appearance parts in loss assessment video |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114462553A (en) * | 2022-04-12 | 2022-05-10 | 之江实验室 | Image labeling and element extraction method and system for car insurance fraud prevention |
| CN114462553B (en) * | 2022-04-12 | 2022-07-15 | 之江实验室 | Image labeling and element extracting method and system for car insurance anti-fraud |
| CN115810134A (en) * | 2023-02-14 | 2023-03-17 | 之江实验室 | Image acquisition quality inspection method, system and device for preventing car insurance from cheating |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10984293B2 (en) | Image processing method and apparatus | |
| CN110766033B (en) | Image processing method, image processing device, electronic equipment and storage medium | |
| CN110570358A (en) | vehicle loss image enhancement method and device based on GAN network | |
| EP3249606B1 (en) | Inverse tone mapping method and corresponding device | |
| CN113706513A (en) | Vehicle damage image analysis method, device, equipment and medium based on image detection | |
| CN109325429A (en) | A kind of method, apparatus, storage medium and the terminal of linked character data | |
| CN111311540A (en) | Vehicle damage assessment method and device, computer equipment and storage medium | |
| CN111415336B (en) | Image tampering identification method, device, server and storage medium | |
| CN110378254B (en) | Method and system for identifying vehicle damage image modification trace, electronic device and storage medium | |
| CN110310247A (en) | Image processing method, device, terminal and computer readable storage medium | |
| CN108323209A (en) | Information processing method, system, cloud processing device and computer program product | |
| US11120308B2 (en) | Vehicle damage detection method based on image analysis, electronic device and storage medium | |
| CN113420871B (en) | Image quality evaluation method, image quality evaluation device, storage medium, and electronic device | |
| CN110909731A (en) | Vehicle rescue scene quality inspection method, device, equipment and storage medium | |
| CN113792600A (en) | Video frame extraction method and system based on deep learning | |
| CN111652145A (en) | Formula detection method and device, electronic equipment and storage medium | |
| US7519237B2 (en) | Method for characterizing stored information | |
| CN115375965A (en) | Preprocessing method for target scene recognition and target scene recognition method | |
| CN112581001B (en) | Evaluation method and device of equipment, electronic equipment and readable storage medium | |
| CN110008940B (en) | Method and device for removing target object in image and electronic equipment | |
| CN114708582A (en) | AI and RPA-based intelligent electric power data inspection method and device | |
| CN111127327B (en) | Picture inclination detection method and device | |
| CN113870210A (en) | Image quality evaluation method, device, equipment and storage medium | |
| CN110889454A (en) | Model training method and device, emotion recognition method and device, equipment and medium | |
| CN113592941B (en) | Certificate image verification method and device, storage medium and terminal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |