Examination registration image clothing detection method and system

Abstract

The embodiment of the present invention relates to the field of image processing technology, and specifically discloses a method and system for detecting clothing in an examination registration image, including: obtaining an image of a candidate, identifying a portrait area from the image of the candidate, identifying a clothing area from the portrait area, and judging whether the clothing area meets a preset condition. If so, determining that the candidate's clothing in the image of the candidate is qualified, and the image of the candidate is qualified; if not, determining that the candidate's clothing in the image of the candidate is unqualified, and determining that the image of the candidate is unqualified. There is no need to manually check the clothing of the candidate's image, and the candidate's clothing can be checked when the candidate's image is obtained, so that the candidate's clothing in the image of the candidate entering the admission ticket printing stage meets the standard requirements, avoiding manual inspection or the need to re-upload the image when the admission ticket is printed because the candidate's clothing in the image of the candidate does not meet the standard, saving a lot of manpower and material resources, and improving the standardization of information in the education examination information system.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and in particular relates to a method and system for detecting clothing in examination registration images. Background Art

[0002] Amid the upsurge in the construction of educational informatization, various educational institutions have built corresponding informatization systems for various educational examinations and admissions, and the collection of registration photos is a very important link.

[0003] Due to the requirements of educational examinations, there are standard requirements for the photos uploaded by candidates who register for the examination, such as specific background color, upright sitting posture, appropriate clothing, etc. However, in the existing technology, the collected photos are usually detected for facial posture and the proportion of the portrait area in the photo, but whether the candidate's clothing is appropriate is not detected. For example, if the candidate is photographed in pajamas, suspenders, etc., or a boy is photographed with his upper body exposed, the uploaded photos will not detect the candidate's clothing, and a large amount of manpower is required to screen the clothing, or the problem of the candidate's clothing in the photo is discovered when printing the admission ticket, and the candidate is required to upload the photo again, wasting a lot of manpower and material resources. Summary of the invention

[0004] The purpose of the embodiments of the present invention is to provide a method and system for detecting clothing in an examination registration image, aiming to solve the problem that the existing technology cannot detect the clothing of the examinee in the photo.

[0005] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions: A method for detecting clothing in an exam registration image, comprising the following steps: Obtaining the examinee's image; Identify a portrait area from the candidate image; identifying a clothing area from the portrait area; Determining whether the dressing area meets the preset conditions; If so, it is determined that the candidate in the candidate image is dressed appropriately, and the candidate image is acceptable; If not, it is determined that the candidate's attire in the candidate image is unqualified and the candidate image is unqualified.

[0006] As a further technical solution of the present invention, obtaining the examinee image specifically includes the following steps: An image uploaded by the examinee is read as the examinee image, or a video uploaded by the examinee is read and an image is captured from the video as the examinee image.

[0007] As a further technical solution of the present invention, identifying a portrait area from the examinee image specifically includes the following steps: Inputting the candidate image into a face recognition model to obtain a face recognition result; Determining whether there is a face in the examinee image based on the face recognition result; If not, a prompt message is generated that the candidate's image does not have a face; If yes, input the candidate image into the portrait recognition model to obtain a portrait detection frame; The portrait detection frame area is segmented from the examinee image as the portrait area.

[0008] As a further technical solution of the present invention, identifying the clothing area from the portrait area specifically includes the following steps: Inputting the portrait area into a face recognition model to obtain a head area; The area below the head area is segmented from the portrait area as a clothing area.

[0009] As a further technical solution of the present invention, judging whether the dressing area meets the preset conditions specifically includes the following steps: Detecting a skin area in the clothing area, and calculating the total area and distribution data of the skin area; Calculating the ratio of the total area of ​​the skin region to the area of ​​the clothing region to obtain the skin ratio; Determining whether the skin ratio is less than a preset ratio; If not, it is determined that the preset conditions are not met; If so, determining whether the skin area is continuous based on the distribution data; If the skin area is continuous, it is determined that the preset condition is met; If the skin area is discontinuous, it is determined that the preset condition is not satisfied.

[0010] As a further technical solution of the present invention, detecting the skin area in the clothing area specifically includes the following steps: Traversing each pixel in the clothing area, and determining whether the pixel value of the current pixel traversed is within a preset pixel value range; If so, the traversed pixel is determined to be a skin pixel; An area consisting of pixels belonging to the skin is determined as a skin area.

[0011] As a further technical solution of the present invention, calculating the total area and distribution data of the skin area specifically includes the following steps: Calculating the number of pixels in the skin area as the total area of ​​the skin area; The number of skin areas in the dressed area is counted.

[0012] As a further technical solution of the present invention, judging whether the skin area is continuous based on the distribution data specifically includes the following steps: Determining whether the number of the skin areas is greater than a preset number; If so, determining that the skin region is continuous; If not, it is determined that the skin region is discontinuous.

[0013] A system for detecting clothing for examination registration images, specifically comprising the following units: An image acquisition unit, used for acquiring an image of the examinee; A portrait region recognition unit, used to recognize a portrait region from the examinee image; a clothing region recognition unit, configured to recognize a clothing region from the portrait region; A judging unit, used for judging whether the dressing area meets the preset condition, and if so, executing the first determining unit, and if not, executing the second determining unit; The first determining unit is used to determine that the examinee in the examinee image is dressed appropriately, and the examinee image is qualified; The second determining unit is used to determine that the dress of the examinee in the examinee image is unqualified, and determine that the examinee image is unqualified.

[0014] As a further technical solution of the present invention, the portrait area recognition unit specifically includes the following modules: A face recognition module, used for inputting the candidate image into a face recognition model to obtain a face recognition result; A face judgment module is used to judge whether there is a face in the candidate image through the face recognition result, and if so, execute the faceless prompt information generation module; if not, execute the portrait detection module; A no-face prompt information generation module, used to generate prompt information that the examinee's image has no face; A portrait detection module, used for inputting the candidate image into a portrait recognition model to obtain a portrait detection frame; The portrait area segmentation module segments the portrait detection frame area from the candidate image as the portrait area.

[0015] Compared with the prior art, the present invention has the following beneficial effects: After acquiring the candidate's image, the embodiment of the present invention identifies a portrait area from the candidate's image, further identifies a clothing area from the portrait area, and determines whether the clothing area meets the preset conditions. If so, it is determined that the candidate's image is qualified; if not, it is determined that the candidate's image is unqualified. This realizes the judgment of whether the preset conditions are met after the clothing area is identified from the candidate's image, and there is no need for manual clothing inspection of the candidate's image. The candidate's clothing can be inspected when the candidate's image is obtained, so that the candidate's clothing in the candidate's image entering the admission ticket printing stage meets the standard requirements, avoids manual inspection of the candidate's clothing in the candidate's image or discovering that the candidate's clothing in the candidate's image does not meet the specified standard requirements when printing the admission ticket, saves a lot of manpower and material resources, and ensures that the candidate's clothing in the uploaded candidate image meets the specified requirements, thereby improving the standardization of information in the education examination information system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention.

[0017] Figure 1 A flow chart of a method for detecting clothing in an examination registration image provided by an embodiment of the present invention is shown.

[0018] Figure 2 Schematic diagram of segmenting the portrait area from the candidate's image.

[0019] Figure 3 A schematic diagram of segmenting the clothing area from the portrait area.

[0020] Figure 4 A flow chart of a method for detecting clothing in an exam registration image in an example of the present invention is shown.

[0021] Figure 5 The application architecture diagram of the examination registration image clothing detection system provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0023] Figure 1 The flowchart of the examination registration image clothing detection method provided by the embodiment of the present invention is shown. Specifically, the examination registration image clothing detection method of the embodiment of the present invention specifically includes the following steps: Step S101, obtaining the examinee's image.

[0024] The examination registration image attire detection method of this embodiment can be integrated into the examination information system to detect the registration image or video uploaded by the examinee to determine whether the examinee's attire in the examinee's image meets the requirements.

[0025] Among them, if the candidate uploads an image when registering, the image uploaded by the candidate can be directly read as the candidate image. If the candidate uploads a video, the video uploaded by the candidate can be read first, and the image can be captured from the video as the candidate image, such as capturing the key frame (I frame) video image as the candidate image.

[0026] Step S102, identifying the portrait area from the candidate's image.

[0027] In an optional embodiment, a face recognition model and a portrait recognition model may be pre-trained, wherein the face recognition model may be a neural network model for performing face detection on an image, which outputs a face recognition result of whether a face exists in the image, and the portrait recognition model may be a neural network model for detecting a portrait area in an image, which outputs the minimum enclosing rectangle of the portrait area in the image as a detection frame.

[0028] Among them, the face recognition model can be trained with a large number of images with labeled faces. For example, a large number of images with labeled faces can be input into the initialized face recognition model to obtain face detection results, and the face detection results are compared with the labeled faces to calculate the loss rate, and then the face recognition model is adjusted by the loss rate until the loss rate is less than a threshold. Among them, the loss rate can be calculated by a mean square error loss function. For example, the first face detection frame is labeled and the second face detection frame is output. The difference in the center points of the first face detection frame and the second face detection frame, the intersection over union (IOU) and other data can be calculated to calculate the loss rate.

[0029] Similarly, when training a portrait recognition model, a large number of labeled portrait areas (portrait annotation boxes) can also be input into the portrait recognition model to obtain a portrait detection box. Then, the loss rate is calculated through the portrait detection box and the portrait annotation box, and the parameters of the portrait recognition model are adjusted until the loss rate is less than the threshold.

[0030] After the face recognition model and the portrait recognition model are trained and integrated into the examination information system, the candidate's image can be input into the face recognition model to obtain the face recognition result, and the face recognition result can be used to determine whether there is a face in the candidate's image. If not, it means that the candidate's facial posture was incorrect when the candidate's image was collected, and the candidate's face cannot be collected. A prompt message indicating that there is no face in the candidate's image can be generated. If so, the candidate's image is further input into the portrait recognition model to obtain a portrait detection frame, and the portrait detection frame area is segmented from the candidate's image as the portrait area.

[0031] like Figure 2 The figure shows a schematic diagram of segmenting the portrait area. After recognizing the face from the candidate image a (the face detection frame in the image b), the portrait detection frame is further recognized (the portrait detection frame in the image c), and then the portrait detection frame area is segmented to obtain the image d, which is the portrait area.

[0032] Step S103, identifying the clothing area from the portrait area.

[0033] In one embodiment, the portrait region may be input into a face recognition model to obtain a head region, and the region below the head region may be segmented from the portrait region as a clothing region.

[0034] like Figure 3 As shown, after the portrait area is input into the face recognition model, the head area is obtained, and the part below the head area is intercepted as the clothing area.

[0035] Step S104, determining whether the clothing area meets the preset conditions.

[0036] In one embodiment, the preset condition may be that the skin ratio is less than a preset ratio and the skin area is continuous. The skin ratio is less than the preset ratio, which constrains the examinee not to expose the upper body or wear clothes that expose too much skin (such as pajamas and home clothes) to take images or videos. The skin area continuity constrains the examinee not to wear suspenders, pajamas, etc. to take images or videos. In this embodiment, step S104 may include the following sub-steps: S1. Detect the skin area in the clothing area and calculate the total area and distribution data of the skin area.

[0037] In one embodiment, the color space of the clothing area can be converted into the YCrCb color space, and then each pixel in the clothing area is traversed, and it is determined whether the pixel value of the current traversed pixel is within a preset pixel value range (the color range of skin color). If so, the traversed pixel is determined to be a skin pixel, and the area composed of the pixels belonging to the skin is determined to be a skin area. If the traversed pixel is not a skin pixel, the pixel is ignored and the traversal continues to the next pixel.

[0038] In another embodiment, the clothing area may be input into a pre-trained semantic segmentation model to obtain the semantics of each area in the clothing area, and the area with the semantics of skin may be determined as the skin area. For example, a semantic segmentation model may be trained using an image labeled with at least a skin area so that the semantic segmentation model can recognize the skin area in the image, and the skin area may be identified by inputting the clothing area into the semantic segmentation model.

[0039] S2. Calculate the ratio of the total area of ​​the skin region to the area of ​​the clothing region to obtain the skin ratio.

[0040] In one embodiment, the area can be represented by the number of pixels. The number of pixels in the skin area can be calculated as the total area of ​​the skin area, and then the total pixels in the clothing area can be calculated as the total area of ​​the clothing area. The ratio of the total area of ​​the skin area to the area of ​​the clothing area is calculated to obtain the skin ratio.

[0041] It should be noted that the clothing area in this embodiment may refer to an area formed by the outline of the human body after removing the background.

[0042] S3. Determine whether the skin ratio is less than a preset ratio.

[0043] This embodiment can pre-configure a reasonable threshold value of the skin ratio, that is, a preset ratio. In one embodiment, a sample of examinees with qualified clothing can be used to perform data statistics to obtain a reasonable threshold value, and then the calculated skin ratio is compared with the preset ratio. If the skin ratio is less than the preset ratio, S4 can be executed. If the skin ratio is greater than or equal to the preset ratio, it means that the examinee's clothing area exposes more skin, which may be due to the examinee wearing suspenders, unbuttoned zippers, pajamas and home clothes, or even taking photos with the upper body exposed, and S6 can be executed.

[0044] S4. Determine whether the skin area is continuous based on the distribution data.

[0045] When the skin ratio is less than the preset ratio, it means that there is little exposed skin in the candidate's image. We can further determine whether the skin is continuous to determine whether the candidate is wearing irregular clothing such as suspenders. Specifically, the number of skin areas in the clothing area can be counted as distribution data to determine whether the number of skin areas is greater than a preset number (if short sleeves are not allowed, the threshold number may be 1; if short sleeves are allowed, the preset number may be 3).

[0046] S5. Determine whether the preset conditions are met.

[0047] If the number of skin areas is less than the preset number, for example, when short sleeves and suspenders are not allowed, the exposed skin is usually the neck area. If the number of skin areas is less than 1, it means that the clothes worn by the examinee cover the shoulders and arms and only the neck area is exposed, which meets the preset conditions. The examinee's clothing is qualified and S105 can be executed.

[0048] S6. Determine whether the preset condition is not satisfied.

[0049] If the number of skin areas is less than the preset number, such as when short sleeves and suspenders are not allowed, the exposed skin is usually the neck area. If the number of skin areas is greater than 1, such as Figure 3As shown, it means that the shoulder area of ​​the examinee is divided into two areas when wearing suspenders and the shoulders and arms are exposed due to wearing short sleeves. It can be determined that the preset conditions are not met, the examinee's dress is unqualified, and S106 can be executed.

[0050] Step S105, determining that the candidate in the candidate image is properly dressed and the candidate image is qualified.

[0051] If the proportion of the skin area in the clothing area is less than a preset ratio, and the skin continuity judges that the skin is continuous, it can be determined that the clothing of the candidate in the candidate image is qualified and the candidate image is also qualified.

[0052] Step S106, determining that the candidate's attire in the candidate image is unqualified, and determining that the candidate image is unqualified.

[0053] If the proportion of the skin area in the clothing area is greater than or equal to a preset ratio, or the skin continuity is judged to be discontinuous, it can be determined that the clothing of the candidate in the candidate's image is unqualified and the candidate's image is also unqualified.

[0054] In order to make those skilled in the art more clear about the process of the examination registration image clothing detection method according to the embodiment of the present invention, the following is combined with Figure 4 The dress inspection process is described as follows: Read photo or video stream information, perform face recognition on the photo or video stream to determine whether there is a face. If not, there is no face information in the range. If so, perform portrait recognition and then portrait segmentation. Cut out the dressing area from the segmented portrait, and further perform skin color pixel analysis on the dressing area to determine whether the area ratio of the skin area exceeds the threshold. If so, return a prompt message that the dress is not standard. If not, perform skin pixel continuity distribution analysis to determine whether the skin pixels are continuous. If so, output the result of dressing standard. If not, return the result of dressing not standard.

[0055] After acquiring the candidate's image, the embodiment of the present invention identifies a portrait area from the candidate's image, further identifies a clothing area from the portrait area, and determines whether the clothing area meets the preset conditions. If so, it is determined that the candidate's image is qualified; if not, it is determined that the candidate's image is unqualified. This realizes the judgment of whether the preset conditions are met after the clothing area is identified from the candidate's image, and there is no need for manual clothing inspection of the candidate's image. The candidate's clothing can be inspected when the candidate's image is obtained, so that the candidate's clothing in the candidate's image entering the admission ticket printing stage meets the standard requirements, avoids manual inspection of the candidate's clothing in the candidate's image or discovering that the candidate's clothing in the candidate's image does not meet the specified standard requirements when printing the admission ticket, saves a lot of manpower and material resources, and ensures that the candidate's clothing in the uploaded candidate image meets the specified requirements, thereby improving the standardization of information in the education examination information system.

[0056] Figure 5The application architecture diagram of the examination registration image clothing detection system provided by an embodiment of the present invention is shown. The examination registration image clothing detection system of this embodiment specifically includes the following units: An image acquisition unit 501 is used to acquire an image of the examinee; A portrait region recognition unit 502, used to recognize a portrait region from the examinee image; A clothing region recognition unit 503, configured to recognize a clothing region from the portrait region; A judging unit 504 is used to judge whether the clothing area meets the preset conditions, and if so, execute the first determining unit 505, and if not, execute the second determining unit 506; The first determining unit 505 is used to determine that the examinee in the examinee image is dressed appropriately, and the examinee image is qualified; The second determining unit 506 is used to determine that the examinee's attire in the examinee image is unqualified, and determine that the examinee image is unqualified.

[0057] As a further technical solution of the embodiment of the present invention, the image acquisition unit 501 specifically includes the following modules: The image acquisition module is used to read the image uploaded by the examinee as the examinee image, or read the video uploaded by the examinee and capture an image from the video as the examinee image.

[0058] As a further technical solution of the embodiment of the present invention, the portrait area recognition unit 502 specifically includes the following modules: A face recognition module, used for inputting the candidate image into a face recognition model to obtain a face recognition result; A face judgment module is used to judge whether there is a face in the candidate image through the face recognition result, and if so, execute the faceless prompt information generation module; if not, execute the portrait detection module; A no-face prompt information generation module, used to generate prompt information that the examinee's image has no face; A portrait detection module, used for inputting the candidate image into a portrait recognition model to obtain a portrait detection frame; The portrait area segmentation module segments the portrait detection frame area from the candidate image as the portrait area.

[0059] As a further technical solution of the embodiment of the present invention, the clothing area identification unit 503 specifically includes the following modules: A head recognition module, used for inputting the portrait area into a face recognition model to obtain a head area; The clothing region segmentation module is used to segment the region below the head region in the portrait region as the clothing region.

[0060] As a further technical solution of the embodiment of the present invention, the judgment unit 504 specifically includes the following modules: A skin detection module, used to detect a skin area in the clothing area and calculate the total area and distribution data of the skin area; A skin ratio calculation module, used to calculate the ratio of the total area of ​​the skin region to the area of ​​the clothing region to obtain the skin ratio; A proportion judgment module is used to judge whether the skin proportion is less than a preset ratio; if so, the continuity judgment module is executed; if not, the condition determination module is not satisfied; A non-satisfied condition determination module, used to determine whether a preset condition is not satisfied; A continuity judgment module, used for judging whether the skin area is continuous based on the distribution data, and if so, executing a module for determining whether a preset condition is satisfied, and if not, executing a module for determining whether a condition is not satisfied; The module for determining whether a preset condition is satisfied is used to determine whether a preset condition is satisfied.

[0061] As a further technical solution of the embodiment of the present invention, the skin detection module specifically includes the following submodules: A pixel traversal and judgment submodule, used to traverse each pixel in the clothing area and judge whether the pixel value of the current pixel traversed is within a preset pixel value range, and if so, execute a skin pixel determination submodule; A skin pixel determination submodule, used to determine the traversed pixels as skin pixels; The skin region determination submodule is used to determine the region composed of pixels belonging to the skin as the skin region.

[0062] As a further technical solution of the embodiment of the present invention, the skin detection module specifically includes the following submodules: A skin area calculation submodule, used for calculating the number of pixels in the skin area as the total area of ​​the skin area; The skin area quantity calculation submodule is used to calculate the quantity of skin areas in the clothing area.

[0063] As a further technical solution of the embodiment of the present invention, the continuity judgment module includes: A skin area quantity determination submodule, used to determine whether the number of the skin areas is greater than a preset number, and if so, execute the first determination submodule, and if not, execute the second determination submodule; A first determination submodule is used to determine that the skin area is continuous; The second determination submodule is used to determine that the skin area is discontinuous.

[0064] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0065] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0066] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0067] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the attached claims.

[0068] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting clothing in an exam registration image, characterized in that: The specific steps include: Obtaining the examinee's image; Identify a portrait area from the candidate image; identifying a clothing area from the portrait area; Determining whether the dressing area meets the preset conditions; If so, it is determined that the candidate in the candidate image is dressed appropriately, and the candidate image is acceptable; If not, it is determined that the candidate's attire in the candidate image is unqualified, and the candidate image is determined to be unqualified.

2. The method for detecting clothing in an examination registration image according to claim 1, characterized in that: Obtaining the candidate image includes the following steps: An image uploaded by the examinee is read as the examinee image, or a video uploaded by the examinee is read and an image is captured from the video as the examinee image.

3. The method for detecting clothing in an examination registration image according to claim 1, characterized in that: Identifying a portrait area from the candidate image specifically includes the following steps: Inputting the candidate image into a face recognition model to obtain a face recognition result; Determining whether there is a face in the examinee image based on the face recognition result; If not, a prompt message is generated that the candidate's image does not have a face; If yes, input the candidate image into the portrait recognition model to obtain a portrait detection frame; The portrait detection frame area is segmented from the examinee image as the portrait area.

4. The method for detecting clothing in an examination registration image according to claim 1, characterized in that: Identifying a clothing area from the portrait area specifically includes the following steps: Inputting the portrait area into a face recognition model to obtain a head area; The area below the head area is segmented from the portrait area as a clothing area.

5. The method for detecting clothing in an examination registration image according to any one of claims 1 to 4, characterized in that: Determining whether the dressing area meets the preset conditions specifically includes the following steps: Detecting a skin area in the clothing area, and calculating the total area and distribution data of the skin area; Calculating the ratio of the total area of ​​the skin region to the area of ​​the clothing region to obtain the skin ratio; Determining whether the skin ratio is less than a preset ratio; If not, it is determined that the preset conditions are not met; If so, determining whether the skin area is continuous based on the distribution data; If the skin area is continuous, it is determined that the preset condition is met; If the skin area is discontinuous, it is determined that the preset condition is not satisfied.

6. The method for detecting clothing in an examination registration image according to claim 5, characterized in that: Detecting the skin area in the clothing area specifically includes the following steps: Traversing each pixel in the clothing area, and determining whether the pixel value of the current pixel traversed is within a preset pixel value range; If so, the traversed pixel is determined to be a skin pixel; An area consisting of pixels belonging to the skin is determined as a skin area.

7. The method for detecting clothing in an examination registration image according to claim 5, characterized in that: Calculating the total area and distribution data of the skin region specifically comprises the following steps: Calculating the number of pixels in the skin area as the total area of ​​the skin area; The amount of skin area in the clothing area is counted.

8. The method for detecting clothing in an examination registration image according to claim 7, characterized in that: Judging whether the skin area is continuous based on the distribution data specifically includes the following steps: Determining whether the number of the skin areas is greater than a preset number; If so, determining that the skin region is continuous; If not, it is determined that the skin region is discontinuous.

9. A system for detecting clothing in examination registration images, characterized in that: It includes the following units: An image acquisition unit, used for acquiring an image of the examinee; A portrait region recognition unit, used to recognize a portrait region from the examinee image; a clothing region recognition unit, configured to recognize a clothing region from the portrait region; A judging unit, used to judge whether the dressing area meets the preset conditions, if yes, execute the first determining unit, if no, execute the second determining unit; The first determining unit is used to determine that the examinee in the examinee image is dressed appropriately, and the examinee image is qualified; The second determining unit is used to determine that the dress of the examinee in the examinee image is unqualified, and determine that the examinee image is unqualified.

10. The examination registration image clothing detection system according to claim 9, characterized in that: The portrait area recognition unit specifically includes the following modules: A face recognition module, used for inputting the candidate image into a face recognition model to obtain a face recognition result; A face judgment module is used to judge whether there is a face in the candidate image through the face recognition result, and if so, execute the faceless prompt information generation module; if not, execute the portrait detection module; A no-face prompt information generation module, used to generate prompt information that the examinee's image has no face; A portrait detection module, used for inputting the candidate image into a portrait recognition model to obtain a portrait detection frame; The portrait area segmentation module segments the portrait detection frame area from the candidate image as the portrait area.

Images