A method, device and apparatus for identifying the judgement area in a court trial video screen

By identifying human body characteristics and judge characteristics on the trial video, combining the legal platform area, the problems of large calculation, high cost and poor adaptability of the authenticity platform area in the prior art are solved, and an efficient and economical legal platform area is achieved. Identify.

CN114241521BActive Publication Date: 2024-11-29CHINASYS TECH
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202111516621.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-13
Publication Date
2024-11-29
Estimated Expiration
2041-12-13

AI Technical Summary

Technical Problem

The prior art has a large calculation workload, high recognition cost when identifying the French station area in the trial video screen, and poor adaptability of the method, which requires manual drawing or matching static items one by one, resulting in low efficiency and high cost.

Method used

By identifying the human body characteristics of the trial video, a person area is obtained, and then the judicial table area is further identified based on the judge's predetermined characteristics, and the judicial table area is obtained in combination, reducing the calculation amount and improving the recognition efficiency, which is suitable for video images of different courts.

Benefits of technology

It reduces the calculation amount of the identification platform area, saves identification costs, improves identification efficiency, and enhances adaptability, and is suitable for a variety of court environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114241521B_ABST
    Figure CN114241521B_ABST
Patent Text Reader

Abstract

This article relates to the field of image recognition, and in particular to a method, device and equipment for identifying the judge's desk area in a court trial video screen. The method includes performing a first feature recognition on the court trial video screen to obtain a human area, wherein the first feature is a human body feature; performing a second feature recognition on the human area to obtain a judge's desk sub-area, wherein the second feature is a predetermined feature of the judge; and combining the judge's desk sub-areas to obtain the judge's desk area in the court trial video screen. Through the embodiments of this article, the recognition of each stationary object in the entire court trial video screen is avoided, and the amount of calculation for detecting the court trial video screen is reduced, thereby improving the efficiency of detecting the judge's desk area in the court trial video screen and saving the recognition cost. In addition, according to the characteristic that the predetermined features of judges in different courts are the same, the method described in the embodiments of this article is applicable to the recognition of the judge's desk area in all court trial video screens, thereby improving adaptability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of image recognition, and more particularly to a method, device and apparatus for identifying a judgement area in a court trial video. Background Art

[0002] During the digital court trial, the behavior of the judges needs to be detected. However, if the entire trial video screen is detected, it will bring huge computational complexity and increase the detection cost. Therefore, the existing technology narrows the detection area by drawing the judge's podium area in the trial video screen to supervise the judge's behavior to prevent the judge from being unfair. For multiple trials, human identification has subjective factors and the area size is not uniform. However, most of the current methods draw a fixed judge's podium area in the trial video screen manually, which is a huge workload.

[0003] The existing technology can also find static landmarks in the entire trial video and match them one by one to finally obtain the judgement area. However, this method requires automatically identifying landmarks in the judgement area in each video image of the trial video, such as the judge's chair, water cup, gavel, etc., which greatly increases the recognition cost. In addition, the size of different courts and the location of cameras in the courts may be different. Therefore, if the judgement area is manually drawn in the trial video of each court, the workload is huge.

[0004] There is an urgent need for a method to identify the judge's desk area in a court trial video screen, so as to solve the problems of large computational workload, high recognition cost and poor adaptability of the recognition method in the prior art for identifying the judge's desk area in a court trial video screen. Summary of the invention

[0005] In order to solve the problems in the prior art of identifying the judgement table area in the court trial video screen, such as large computational workload, high recognition cost, and poor adaptability of the recognition method, the embodiments of this article provide a method, device and equipment for identifying the judgement table area in the court trial video screen, which reduces the computational workload of detecting the court trial video screen, thereby improving the efficiency of detecting the judgement table area in the court trial video screen, saving recognition cost, and enhancing adaptability.

[0006] In order to solve the above technical problems, the specific technical solutions of this article are as follows:

[0007] On the one hand, the embodiment of this document provides a method for identifying a judgement area in a court trial video screen, comprising:

[0008] Performing first feature recognition on the court trial video screen to obtain a human area, wherein the first feature is a human body feature;

[0009] Performing second feature recognition on the manned area to obtain a judge's bench sub-area, wherein the second feature is a predetermined feature of the judge;

[0010] The judge platform sub-areas are combined to obtain the judge platform area in the trial video screen.

[0011] Furthermore, before performing the first feature recognition on the court trial video screen to obtain the human area, it also includes:

[0012] Detecting moving objects in the trial video screen to obtain a moving object area;

[0013] Calculating the proportion of the moving object area relative to the trial video screen;

[0014] Removing the area whose proportion is greater than a first preset threshold value;

[0015] Performing a first feature recognition on the court trial video screen to obtain a human area further includes: performing a first feature recognition on the moving object area after removing the area with a proportion greater than a first preset threshold value to obtain a human area.

[0016] Furthermore, after performing a second feature recognition on the human area to obtain the platform sub-area, the method further includes:

[0017] Select one of the French platform sub-areas as the lateral reference area;

[0018] According to a preset step length, horizontally translate the boundary box of the horizontal reference area in the trial video screen to obtain a suspected judge platform sub-area;

[0019] The second feature recognition is performed on the suspected law platform sub-region, and the suspected law platform sub-region in which the second feature is recognized is used as the law platform sub-region.

[0020] Furthermore, the lateral reference area is adjacent to only one bench sub-area in the horizontal direction.

[0021] Furthermore, before determining the lateral reference area in the base sub-area, the method further includes:

[0022] The judge platform sub-area close to the bottom edge of the trial video screen is used as the vertical reference area;

[0023] Calculating the longitudinal distance between the bench sub-area excluding the longitudinal reference area and the longitudinal reference area;

[0024] The platform sub-area whose longitudinal distance exceeds a second preset threshold value is removed.

[0025] Furthermore, performing second feature recognition on the human area further includes:

[0026] If the second feature is not recognized in the human-inhabited area, expanding the human-inhabited area according to a predetermined range;

[0027] A second feature recognition is performed on the enlarged human-containing area.

[0028] Furthermore, the judge platform sub-areas are combined to obtain the judge platform area in the trial video screen, further including:

[0029] Filtering out the boundary closest to the boundary of the trial video screen from the boundary of the judge platform sub-area;

[0030] The area enclosed by the extended line of the screened boundary is used as the legal platform area.

[0031] On the other hand, the embodiment of this invention also provides a device for identifying the judge's platform area in the court trial video screen, including:

[0032] A human region determination unit is used to perform first feature recognition on the trial video screen to obtain a human region, wherein the first feature is a human body feature;

[0033] A judge's bench sub-area determination unit, configured to perform second feature recognition on the manned area to obtain a judge's bench sub-area, wherein the second feature is a predetermined feature of the judge;

[0034] The judgement platform area determination unit is used to combine the judgement platform sub-areas to obtain the judgement platform area in the trial video screen.

[0035] On the other hand, an embodiment of the present invention further provides a computer device, including a memory, a processor, and a computer program stored in the memory, and the processor implements the above method when executing the computer program.

[0036] Finally, the embodiments of this document also provide a computer storage medium on which a computer program is stored. When the computer program is executed by a processor of a computer device, the above method is executed.

[0037] Using the embodiments of this article, because the number of people in the trial video screen is far less than the number of static objects, the human body features are first identified, avoiding the identification of each static object in the entire trial video screen. Since the features of the judge in the trial video screen are more obvious than the features of other people in the trial video screen, the identified area with people is further identified according to the judge's predetermined features to obtain the judge's bench sub-area, and finally the bench sub-area is combined to obtain the bench area, which reduces the amount of calculation for detecting the trial video screen, thereby improving the efficiency of detecting the bench area in the trial video screen and saving the identification cost. In addition, according to the feature that the predetermined features of judges in different courts are the same, the method described in the embodiments of this article is applicable to the identification of the bench area in all trial video screens, which improves adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of this article or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of this article. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 Shown is a schematic diagram of the trial video screen of the embodiment of this article;

[0040] Figure 2 It is a flow chart of a method for identifying the judge's platform area in a court trial video screen according to an embodiment of this invention;

[0041] Figure 3 The figure shows the process of removing the incentive picture from the trial video picture in the embodiment of this article;

[0042] Figure 4 The figure shows the process of determining the suspected law platform sub-region in the embodiment of this article;

[0043] Figure 5 The figure shows the process of filtering the sub-areas of the bench in the vertical direction according to the embodiment of this article;

[0044] Figure 6 The figure shows the process of expanding the occupied area according to the predetermined range in the embodiment of this article;

[0045] Figure 7 The figure shows a schematic diagram of the structure of a device for identifying the judge's platform area in a court trial video screen according to an embodiment of this invention;

[0046] Figure 8 The figure is a schematic diagram of the process of identifying the judge's platform area in the court trial video screen according to the embodiment of this article;

[0047] Fig. 9 Shown is a schematic diagram of the structure of the computer device of the embodiment of this article.

[0048]

Description of reference numerals

[0049] 101. Video footage of the trial;

[0050] 102. French platform area;

[0051] 701, Manned Area Determination Unit;

[0052] 702, French platform sub-region determination unit;

[0053] 703, French and Taiwanese regional determination unit;

[0054] 902. Computer equipment;

[0055] 904. Processing equipment;

[0056] 906. Storage resources;

[0057] 908, driving mechanism;

[0058] 910, input / output module;

[0059] 912. Input device;

[0060] 914. Output device;

[0061] 916. Presentation equipment;

[0062] 918. Graphical user interface;

[0063] 920, network interface;

[0064] 922, communication link;

[0065] 924. Communication bus. DETAILED DESCRIPTION

[0066] The following will be combined with the drawings in the embodiments of this article to clearly and completely describe the technical solutions in the embodiments of this article. Obviously, the described embodiments are only part of the embodiments of this article, not all of the embodiments. Based on the embodiments of this article, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this article.

[0067] It should be noted that the terms "first", "second", etc. in the specification and claims of this article and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of this article described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0068] like Figure 1 The figure shows a schematic diagram of the trial video screen of the embodiment of this article, which includes the trial video screen 101 and the judge area 102. It should be noted that: Figure 1The trial video screen shown is only for illustration and does not represent the trial video screen in the actual trial scene. The trial video screen 101 described in the embodiment of this article can be composed of multiple video screens, each of which is recorded by a camera set in the courtroom according to a specified recording angle. The trial video screen includes the judge's area, the plaintiff's area, the defendant's area, the audience seats, etc., and the judge's area 102 is located in the judge's area. In addition, Figure 1 The trial video screen shown in the trial video screen 101 can also include an incentive screen, which is an enlarged video screen of the judge's area, the plaintiff's area or the defendant's area. The video recording system detects which area is speaking, enlarges the area, and obtains the incentive screen. Therefore, the incentive screen changes with the change of the speaker.

[0069] In order to supervise the behavior of the judge, it is necessary to separately identify the judge's bench area 102. The existing technology requires manual recognition of the trial video screen 101 and drawing the bench area 102. Due to the involvement of human subjective factors, the accuracy of identifying the bench area 102 is low. If the static landmark objects in the trial video screen 101 are matched one by one to obtain the bench area 102, the recognition cost is greatly increased. In addition, since the size of different courts and the position of the cameras in the courts may be different, it is a huge workload to manually draw the bench area 102 in the trial video screen 101 of each court.

[0070] In order to solve the problem that the existing technology for identifying the judge area in the court trial video screen has a large computational workload, high recognition cost, and poor adaptability of the recognition method, the embodiment of this article provides a method for identifying the judge area in the court trial video screen, which reduces the computational workload of detecting the court trial video screen, thereby improving the efficiency of detecting the judge area in the court trial video screen, saving recognition cost, and enhancing adaptability.

[0071] like Figure 2 The flowchart of a method for identifying the judge's seat area in a court trial video screen according to an embodiment of the present invention is shown. This figure describes the process of identifying the judge's seat area in a court trial video screen, but it may include more or fewer operation steps based on conventional or non-creative work. The order of steps listed in the embodiment is only one way of executing the steps among many orders, and does not represent the only order of execution. When the system or device product is executed in practice, it can be executed in the order or in parallel according to the method shown in the embodiment or the accompanying drawings. Specifically, Figure 2 As shown, the method may include:

[0072] Step 201: performing first feature recognition on the court trial video screen to obtain a human area, wherein the first feature is a human body feature;

[0073] Step 202: performing second feature recognition on the human-occupied area to obtain a judge's bench sub-area, wherein the second feature is a predetermined feature of the judge;

[0074] Step 203: Combine the judge platform sub-areas to obtain the judge platform area in the trial video screen.

[0075] Through the method of the embodiment of this article, because the number of people in the trial video screen is far less than the number of static objects, the human body features are first identified, avoiding the identification of each static object in the entire trial video screen. Since the features of the judge in the trial video screen are more obvious than the features of other people in the trial video screen, the identified area with people is further identified according to the judge's predetermined features to obtain the judge's bench sub-area, and finally the bench sub-area is combined to obtain the bench area, which reduces the amount of calculation for detecting the trial video screen, thereby improving the efficiency of detecting the bench area in the trial video screen and saving the identification cost. In addition, according to the feature that the predetermined features of judges in different courts are the same, the method described in the embodiment of this article is applicable to the identification of the bench area in all trial video screens, which improves adaptability.

[0076] In the embodiments of this article, the first feature is a human body feature, such as the structure of the upper body of the human body, including the head, neck, shoulders, etc. The second feature is a predetermined feature of the judge, such as the robes and judge's chair worn by the judge. The identified judicial sub-area includes the judge's chair, and the judicial sub-area and the judge's chair correspond one to one. The first feature recognition is performed on the trial video screen, and one or more areas with people can be obtained, and each area with people can be represented by a rectangular frame. The second feature recognition is performed on the area with people, and the area with people that recognizes the second feature can be used as the judicial sub-area, and the area with people that is not recognized as having the second feature can be discarded. Finally, the one or more judicial sub-areas obtained are combined to obtain the judicial area, and the combination method can be spliced ​​according to the positional relationship of each judicial sub-area in the trial video screen. It should be noted that the first feature recognition of the trial video screen and the second feature recognition of the area with people can be performed by the image recognition method in the prior art, which is not limited in the embodiments of this article.

[0077] According to an embodiment of this invention, in order to further reduce the amount of calculation and realize the identification of the judicial area of ​​the trial video screen with the incentive screen, as shown in FIG. Figure 3 As shown, step 201 performs a first feature recognition on the court trial video screen to obtain a human area, and also includes:

[0078] Step 301: Detect moving objects in the trial video screen to obtain a moving object area;

[0079] Step 302: Calculate the proportion of the moving object area relative to the trial video screen;

[0080] Step 303: remove the area whose proportion is greater than the first preset threshold value.

[0081] The first feature recognition is performed on the trial video screen, and the human area is further obtained including:

[0082] Step 304: Perform a first feature recognition on the moving object area after removing the area whose proportion is greater than the first preset threshold value to obtain a human area.

[0083] In the embodiments of this invention, since the positions of moving objects in multiple adjacent frames of video are different, the color changes of pixels in multiple adjacent frames of trial video can be detected to obtain a moving object region, which may include moving people, moving plants, etc. Since the detection complexity of a moving object region is much smaller than that of a stationary object, the method for detecting moving objects in a trial video in the embodiments of this invention has a relatively small amount of computation.

[0084] In addition, since the incentive picture in the trial video screen is obtained by enlarging the area where someone is speaking, the proportion of the moving object area relative to the trial video screen can be calculated. For example, the moving object area A in the incentive picture is an enlargement of the area A' where the defendant is located, and the proportion of the moving object area A in the incentive picture relative to the trial video screen is greater than the proportion of the area A' where the defendant is located relative to the trial video screen. Therefore, when the proportion is greater than the first preset threshold value, the picture represented by the proportion is determined to be the incentive picture. In the embodiment of this article, the first preset gate limit can be obtained by analyzing a large number of trial video screens, for example, obtaining the moving object area in multiple trial video screens, and then calculating the size of the moving object area in the incentive picture of each trial video screen and the size in the non-magnified picture, and then respectively calculating the proportion of the two relative to the trial video screen. A large number of trial video screens are analyzed by the above method to determine the first preset threshold value.

[0085] In addition, step 302 also calculates the ratio of the difference between the pixels of the two moving object regions with the same content obtained in the court trial video screen to ...

[0086]

[0087] Wherein, p represents the proportion, N Arepresents the number of pixels of the moving object region with the number of pixels ranked first among the two moving object regions with the same content, N A′ represents the number of pixels of the moving object region whose number of pixels is ranked later in the two moving object regions with the same content, and N represents the total number of pixels of the trial video screen. Preferably, the first preset threshold value is 6‰.

[0088] Then, the first feature recognition is performed on the moving object area after removing the area whose proportion is greater than the first preset threshold value, and the human area is obtained, and the influence of the incentive picture can be removed. The above steps 301 to 304 further reduce the amount of calculation, and remove the influence of the incentive picture in the trial video screen on the identification of the judicial area, thereby improving the accuracy of identifying the judicial area.

[0089] In the judgement area of ​​the trial video screen described in the embodiment of this article, there may be a situation where the judge does not participate in the trial, so it is necessary to detect the judge who does not participate in the trial. Therefore, according to one embodiment of this article, Figure 4 As shown, after step 202 performs second feature recognition on the human area to obtain the platform sub-area, it also includes:

[0090] Step 401: Select one of the bench sub-areas as a lateral reference area;

[0091] Step 402: translating the boundary box of the horizontal reference area horizontally in the court trial video screen according to a preset step length to obtain a suspected judge's platform sub-area;

[0092] Step 403: performing the second feature recognition on the suspected law platform sub-region, and using the suspected law platform sub-region in which the second feature is recognized as the law platform sub-region.

[0093] In the embodiment of this article, the judge's desk sub-area obtained in step 202 is a region with people, and does not include the judge who did not attend the trial. Therefore, it is necessary to select one of the judge's desk sub-areas as the horizontal reference area, for example, randomly select a judge's desk sub-area as the horizontal reference area, and the horizontal reference area is a region with people, that is, there is a judge in the horizontal reference area. Then, according to the preset step length, the boundary box of the selected horizontal reference area is horizontally translated in the trial video screen to obtain the suspected judge's desk area, wherein the preset step length can be determined according to the size of the boundary box of the judge's desk sub-area. As a preferred embodiment, the preset step length described in the embodiment of this article is equal to the size of the horizontal boundary of the judge's desk sub-area, wherein the suspected judge's desk area indicates that the area may include a predetermined feature of the judge, so the suspected judge's desk area is identified with the second feature, and the suspected judge's desk area with the second feature identified is used as the judge's desk sub-area, and the judge's desk sub-area may include the judge, or include the judge's chair with the second predetermined feature but does not include the judge. Therefore, through the above method, the judge who did not attend the trial is detected, and the number of judges who did not attend the trial is obtained, so as to further determine the judges who did not attend the trial according to the trial record.

[0094] It should be noted that, in step 402, multiple suspected law platform areas can be obtained, and then the second feature recognition is performed on each suspected law platform sub-area respectively to obtain one or more law platform sub-areas. In addition, the boundary box of the horizontal reference area can be horizontally translated once to obtain a suspected law platform sub-area, and then the second feature recognition is performed on the suspected law platform sub-area in step 403. If the second feature is not recognized in the suspected law platform sub-area in step 403, it means that the suspected law platform sub-area is already outside the boundary of the law platform area, and the operation of step 402 is stopped to reduce the amount of calculation.

[0095] According to an embodiment of the present invention, in order to improve the efficiency of identifying the suspected law platform sub-region, the lateral reference region described in step 401 is adjacent to only one law platform sub-region in the horizontal direction.

[0096] In the embodiment of this article, the lateral reference area is adjacent to only one bench sub-area in the horizontal direction, that is, the lateral reference area is the leftmost or rightmost bench sub-area in the bench sub-area. Preferably, the lateral reference area described in the embodiment of this article is the leftmost bench sub-area in the bench sub-area. Then, the bounding box of the lateral reference area is horizontally translated to the right according to a preset step length to obtain a suspected bench sub-area on the right side of the lateral reference area, and then the bounding box of the lateral reference area is horizontally translated to the left according to the preset step length to obtain a suspected bench sub-area on the left side of the lateral reference area. Since the lateral reference area is adjacent to only one bench sub-area in the horizontal direction, multiple identification of the same area is avoided, thereby improving the efficiency of identifying suspected bench sub-areas.

[0097] In the judge's desk area of ​​the trial video screen described in the embodiment of this article, there may be judges' chairs that are not on the same horizontal line, which may indicate that the court staff has moved the judge's chair who has requested to leave behind the other judges' chairs. When the above situation occurs, the area where the judge's chair moved behind the other judges' chairs is located in the judge's desk area should be filtered out. Therefore, according to one embodiment of this article, in order to filter out the area where the judge's chair moved behind the other judges' chairs is located, Figure 5 As shown, before determining the lateral reference area in the bench sub-area in step 401, the step further includes:

[0098] Step 501: taking the judge's platform sub-area close to the bottom edge of the trial video screen as a vertical reference area;

[0099] Step 502: Calculate the longitudinal distance between the bench sub-area except the longitudinal reference area and the longitudinal reference area;

[0100] Step 503: Remove the platform sub-area whose longitudinal distance exceeds the second preset threshold value.

[0101] In the embodiment of this article, according to the structure of the bench, the judge appearing in court should be close to the table in the bench area. Therefore, the bench sub-area close to the bottom edge of the trial video screen is first used as the longitudinal reference area, that is, the bottom bench sub-area is used as the longitudinal reference area, and then the longitudinal distance between the remaining bench sub-areas and the longitudinal reference area is calculated. The longitudinal distance can indicate whether each bench sub-area is in the same row as the longitudinal reference area. If some bench sub-areas belong to the same row as the longitudinal reference area, the longitudinal distance between them is small. If some bench sub-areas are not in the same row as the longitudinal reference area, the longitudinal distance between them is large. Therefore, when the longitudinal distance exceeds the second preset threshold value, it means that the bench sub-area to which the longitudinal distance belongs is not in the same row as the longitudinal reference area. It is removed from the bench sub-area, and the area where the judge's chair is placed behind the remaining judges' chairs can be filtered out. In the embodiment of this article, the second preset threshold value can be obtained by analyzing the actual measured data at the trial site. In the actual trial site, a judge may appropriately adjust the position of the judge's chair, so the second preset threshold value is greater than 0.

[0102] In the embodiment of this article, the size of the bounding box of the manned area identified in step 201 may be smaller than the size of the judge's chair. When the above situation occurs, although the second feature may not be identified in the manned area, the manned area does belong to the area where the judge is located. Therefore, according to an embodiment of this article, in order to solve the problem that the size of the bounding box of the manned area is smaller than the size of the judge's chair, resulting in inaccurate identification of the judge's table sub-area, as shown in FIG. Figure 6 As shown, step 202 further includes performing second feature recognition on the human area,

[0103] Step 601: If the second feature is not recognized in the human area, the human area is expanded according to a predetermined range;

[0104] Step 602: Perform a second feature recognition on the enlarged human-containing area.

[0105] In the embodiments of this article, the predetermined range can represent the horizontal and vertical expansion range of the bounding box of the manned area, and the horizontal and vertical expansion range can be obtained based on the analysis of the actual measured data at the trial site. For example, the width and height of the judge's chair are measured, and the horizontal expansion range of the bounding box of the manned area is determined according to the width of the judge's chair, and the vertical expansion range of the bounding box of the manned area is determined according to the height of the judge's chair. Therefore, if the manned area where there is actually a judge does not identify the predetermined features of the judge, the manned area can be expanded to include the predetermined features of the judge, which solves the problem that the size of the bounding box of the manned area is smaller than the size of the judge's chair, resulting in inaccurate identification of the sub-area of ​​the judge's platform.

[0106] In the embodiments of this article, a judge's desk sub-area may be an area where only one judge is located. If multiple staff members are arranged to supervise the judge's behavior, the workload may be large, and each judge's desk sub-area is displayed separately, and it is impossible to associate the judge's desk sub-area. For example, it is impossible to judge the judge's actual intention to lean to one side. Therefore, it is necessary to combine the judge's desk sub-areas to obtain the judge's desk area, so that the staff can supervise the judge's behavior.

[0107] Therefore, according to an embodiment of the present invention, step 203 combines the judge platform sub-areas to obtain the judge platform area in the trial video screen further including:

[0108] Filtering out the boundary closest to the boundary of the trial video screen from the boundary of the judge platform sub-area;

[0109] The area enclosed by the extended line of the screened boundary is used as the legal platform area.

[0110] In the embodiments of this article, the boundary closest to the boundary of the court video screen is screened out from the boundary of the court sub-area, and the area enclosed by the extension line of the screened boundary is used as the court area, that is, a boundary box that can contain all the court sub-areas is obtained in the court video screen, and then the area enclosed by this boundary box is used as the court area, and the court sub-areas can be combined together to facilitate the staff to supervise the behavior of the judge.

[0111] Based on the same inventive concept, the embodiment of this invention also provides a device for identifying the judge's platform area in the court trial video screen. Figure 7As shown, it includes a manned area determination unit 701, a law platform sub-area determination unit 702, and a law platform area determination unit 703:

[0112] A human region determination unit 701 is used to perform first feature recognition on the court trial video screen to obtain a human region, where the first feature is a human body feature;

[0113] A judge's bench sub-region determining unit 702 is used to perform second feature recognition on the manned area to obtain a judge's bench sub-region, where the second feature is a predetermined feature of the judge;

[0114] The judgement table area determination unit 703 is used to combine the judgement table sub-areas to obtain the judgement table area in the trial video screen.

[0115] The beneficial effects obtained by the above-mentioned device are consistent with the beneficial effects obtained by the above-mentioned method, and will not be described in detail in the embodiments of this specification.

[0116] like Figure 8 Shown is a flow chart of identifying the judge desk area in the trial video screen in the embodiment of this invention. This figure describes the steps of identifying the judge desk area in the trial video screen. It should be noted that the steps and sequence described in this figure are not the only steps and sequence for identifying the judge desk area in the trial video screen in the embodiment of this invention. For those skilled in the art, other steps and sequences for identifying the judge desk area in the trial video screen can also be obtained based on the content described in this figure, and this specification embodiment does not limit this.

[0117] Specifically, the steps of identifying the judge's platform area in the court trial video screen include:

[0118] Step 801: Detect moving objects in the trial video screen to obtain a moving object area;

[0119] In this step, since the positions of moving objects in multiple adjacent frames of video are different, the color changes of pixel points in multiple adjacent frames of trial video can be detected to obtain the moving object area, which may include moving people, moving plants, etc.

[0120] Step 802: Calculate the proportion of the moving object area relative to the trial video screen;

[0121] In this step, since the incentive picture in the trial video is obtained by magnifying the area where someone is speaking, it is possible to determine whether the moving object area is the incentive picture by calculating the proportion of the moving object area relative to the trial video picture.

[0122] Step 803: removing the area whose proportion is greater than a first preset threshold value;

[0123] In this step, when the proportion is greater than the first preset threshold value, the picture represented by the proportion is identified as the incentive picture, and the influence of the incentive picture on the identification of the judicial area can be removed by removing the area with a proportion greater than the first preset threshold value. In the embodiment of this article, the first preset gate limit can be obtained by analyzing a large number of trial video pictures.

[0124] Step 804: performing a first feature recognition on the moving object region after removing the region whose proportion is greater than the first preset threshold value to obtain a human-occupied region;

[0125] In this step, the first feature is a human body feature, such as the structure of the upper body, including the head, neck, shoulders, etc. By performing first feature recognition on the court trial video, one or more human areas can be obtained, and each human area can be represented by a rectangular frame.

[0126] Because the number of people in the trial video is far less than the number of static objects, the human body features are identified first, avoiding the identification of every static object in the entire trial video and reducing the amount of calculation for detecting the trial video.

[0127] Step 805: performing a second feature recognition on the human-occupied area to obtain a platform sub-area;

[0128] In this step, since the judge's features in the court trial video are more obvious than the features of other people in the court trial video, the identified human area is further identified according to the judge's predetermined features to obtain the judge's bench sub-area. The second feature may be, for example, the judge's robe, the judge's chair, etc. The identified bench sub-area includes the judge's chair, and the bench sub-area and the judge's chair correspond one to one.

[0129] Step 806: The second feature is not identified;

[0130] In this step, if the second feature is not identified in the human area, it means that the size of the human area boundary box may be smaller than the size of the judge's chair. Although the second feature is not identified, in the actual trial, the human area belongs to the area where the judge is located. Therefore, when the second feature is not identified, step 807 is executed to expand the human area.

[0131] Step 807: Expand the occupied area according to a predetermined range;

[0132] In this step, the predetermined range can represent the horizontal and vertical expansion range of the bounding box of the manned area, and the horizontal and vertical expansion range can be obtained based on the actual measured data analysis of the trial site. For example, the width and height of the judge's chair are measured, and the horizontal expansion range of the bounding box of the manned area is determined according to the width of the judge's chair, and the vertical expansion range of the bounding box of the manned area is determined according to the height of the judge's chair. Therefore, if the manned area where there is actually a judge does not identify the predetermined features of the judge, the manned area is expanded according to the predetermined range, and the expanded manned area can include the predetermined features of the judge (such as the judge's chair), which solves the problem of inaccurate identification of the judge's platform sub-area when the size of the bounding box of the manned area is smaller than the size of the judge's chair. After the manned area is expanded according to the predetermined range, step 805 is executed again to perform a second feature recognition on the expanded manned area.

[0133] It should be noted that the method for identifying the judge's platform area in the trial video screen described in the embodiment of this article only expands the area with people in it where the second feature is not identified once according to a predetermined range. If the second feature is still not identified in the expanded area with people in it, it is considered that the area with people in it is not the area where the judge is located, and the area with people in it will no longer be expanded.

[0134] Step 808: taking the judge's platform sub-area close to the bottom edge of the trial video screen as a vertical reference area;

[0135] In order to filter out the judge chairs that are moved behind the other judge chairs, the bench sub-area needs to be filtered according to the longitudinal reference area. In this step, according to the structure of the bench, the judges in court should be close to the table in the bench area, so the bench sub-area close to the bottom edge of the trial video screen is first used as the longitudinal reference area, that is, the bottom bench sub-area is used as the longitudinal reference area.

[0136] Step 809: Calculate the longitudinal distance between the bench sub-area except the longitudinal reference area and the longitudinal reference area;

[0137] In this step, the longitudinal distances between the remaining law table sub-areas and the longitudinal reference area are calculated. The longitudinal distances can indicate whether each law table sub-area and the longitudinal reference area are in the same row. If some law table sub-areas and the longitudinal reference area are in the same row, the longitudinal distance between them is smaller. If some law table sub-areas and the longitudinal reference area are not in the same row, the longitudinal distance between them is larger.

[0138] Step 810: removing the bench sub-area whose longitudinal distance exceeds a second preset threshold value;

[0139] In this step, when the longitudinal distance exceeds the second preset threshold value, it means that the judge's desk sub-area to which the longitudinal distance belongs is not in the same row as the longitudinal reference area. By removing it from the judge's desk sub-area, the area where the judge's chair that is moved behind the other judges' chairs is located can be filtered out. In the embodiment of this article, the second preset threshold value can be obtained by analyzing the actual measured data of the trial scene. In the actual trial scene, a judge may adjust the position of the judge's chair, so the second preset threshold value is greater than 0.

[0140] Step 811: Selecting a law platform sub-region that is adjacent to only one law platform sub-region in the horizontal direction as a lateral reference region;

[0141] In order to detect the judges who did not attend the trial, the judge platform sub-area needs to be expanded. In this step, the horizontal reference area is the leftmost or rightmost judge platform sub-area in the judge platform sub-area. Preferably, the horizontal reference area described in the embodiment of this article is the leftmost judge platform sub-area in the judge platform sub-area.

[0142] Step 812: translating the boundary box of the horizontal reference area horizontally in the court trial video screen according to a preset step length to obtain a suspected judge's platform sub-area;

[0143] In this step, the boundary box of the horizontal reference area is horizontally translated to the right according to the preset step size to obtain the suspected law platform sub-area on the right side of the horizontal reference area, and then the boundary box of the horizontal reference area is horizontally translated to the left according to the preset step size to obtain the suspected law platform sub-area on the left side of the horizontal reference area.

[0144] Step 813: performing the second feature recognition on the suspected law platform sub-region, and using the suspected law platform sub-region in which the second feature is recognized as the law platform sub-region;

[0145] In this step, the suspected judgement desk area is subjected to second feature recognition, and the suspected judgement desk area with the second feature identified is used as the judgement desk sub-area. The judgement desk sub-area may include the judge, or include the judge's chair with the second predetermined feature but not the judge. Thus, the judges who did not participate in the trial are detected, and the number of judges who did not participate in the trial is obtained, so that the judges who did not participate in the trial can be further determined based on the trial records.

[0146] It should be noted that in step 812, multiple suspected law platform areas can be obtained, and then the second feature recognition is performed on each suspected law platform sub-area to obtain one or more law platform sub-areas. It is also possible to horizontally translate the boundary box of the horizontal reference area once to obtain a suspected law platform sub-area, and then perform the second feature recognition on the suspected law platform sub-area in step 813. If the second feature is not recognized in the suspected law platform sub-area in step 813, it means that the suspected law platform sub-area is already outside the boundary of the law platform area, and the operation of step 812 is stopped to reduce the amount of calculation.

[0147] Step 814: Filter out the boundary closest to the boundary of the trial video screen from the boundary of the judge platform sub-area;

[0148] In this step, in addition to screening out the boundary closest to the boundary of the trial video screen from the boundary of the judge platform sub-area, the coordinates of the intersection points of the boundary lines of the same judge platform sub-area can also be determined, and then the above coordinates of all judge platform areas are compared to select the smallest coordinate value in the longitudinal direction, the smallest coordinate value in the transverse direction, the largest coordinate value in the longitudinal direction, and the largest coordinate value in the transverse direction, so that step 815 can use the area surrounded by these four coordinate values ​​as the judge platform area.

[0149] Step 815: The area enclosed by the extended lines of the filtered boundaries is used as the legal platform area.

[0150] In this step, in addition to using the area enclosed by the extension line of the selected boundary as the bench area, the area enclosed by the smallest coordinate value in the longitudinal direction, the smallest coordinate value in the transverse direction, the largest coordinate value in the longitudinal direction, and the largest coordinate value in the transverse direction obtained in step 814 can also be used as the bench area. Combining the bench sub-areas together makes it easier for staff to supervise the behavior of the judges.

[0151] like Fig. 9The structure diagram of the computer device of the embodiment of this article is shown. The device in this article can be a computer device in this embodiment, and the method of this article is executed. The computer device 902 may include one or more processing devices 904, such as one or more central processing units (CPUs), and each processing unit can implement one or more hardware threads. The computer device 902 may also include any storage resource 906, which is used to store any kind of information such as code, settings, data, etc. Non-limiting, for example, the storage resource 906 may include any one or more combinations of the following: any type of RAM, any type of ROM, flash memory device, hard disk, optical disk, etc. More generally, any storage resource can use any technology to store information. Further, any storage resource can provide volatile or non-volatile retention of information. Further, any storage resource can represent a fixed or removable component of the computer device 902. In one case, when the processing device 904 executes an associated instruction stored in any storage resource or a combination of storage resources, the computer device 902 can perform any operation of the associated instruction. The computer device 902 also includes one or more drive mechanisms 908 for interacting with any storage resources, such as a hard disk drive mechanism, an optical disk drive mechanism, and the like.

[0152] The computer device 902 may also include an input / output module 910 (I / O) for receiving various inputs (via input devices 912) and for providing various outputs (via output devices 914). A specific output mechanism may include a presentation device 916 and an associated graphical user interface (GUI) 918. In other embodiments, the input / output module 910 (I / O), the input device 912, and the output device 914 may not be included, and the computer device 902 may be used as a computer device in a network. The computer device 902 may also include one or more network interfaces 920 for exchanging data with other devices via one or more communication links 922. One or more communication buses 924 couple the components described above together.

[0153] The communication link 922 may be implemented in any manner, for example, through a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication link 922 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc. governed by any protocol or combination of protocols.

[0154] Corresponds to Figure 2-Figure 6 , Figure 8 The method in this article also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above steps are performed.

[0155] The embodiment of the present invention also provides a computer readable instruction, wherein when the processor executes the instruction, the program therein causes the processor to execute the following Figure 2-Figure 6 , Figure 8 The method shown.

[0156] It should be understood that in the various embodiments of this document, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this document.

[0157] It should also be understood that in the embodiments of this article, the term "and / or" is only a description of the association relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0158] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this article.

[0159] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0160] In the several embodiments provided herein, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or can be electrical, mechanical or other forms of connection.

[0161] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of this article.

[0162] In addition, each functional unit in each embodiment of this invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional unit.

[0163] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this article is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of this article. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0164] Specific embodiments are used in this article to illustrate the principles and implementation methods of this article. The description of the above embodiments is only used to help understand the methods and core ideas of this article. At the same time, for general technicians in this field, according to the ideas of this article, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on this article.

Claims

1. A method for identifying the judgement area in a court trial video, characterized in that: The method comprises, Performing first feature recognition on the court trial video screen to obtain a human area, wherein the first feature is a human body feature; Performing second feature recognition on the manned area to obtain a judge's bench sub-area, wherein the second feature is a predetermined feature of the judge; The judge platform sub-areas are combined to obtain the judge platform area in the trial video screen, and the combination method is to splice them according to the position relationship of each judge platform sub-area in the trial video screen.

2. The method for identifying the judgement area in the court trial video according to claim 1, characterized in that: Before the first feature recognition of the court trial video screen is performed to obtain the human area, it also includes: Detecting moving objects in the trial video screen to obtain a moving object area; Calculating the proportion of the moving object area relative to the trial video screen; Removing the area whose proportion is greater than a first preset threshold value; Performing a first feature recognition on the court trial video screen to obtain a human area further includes: performing a first feature recognition on the moving object area after removing the area with a proportion greater than a first preset threshold value to obtain a human area.

3. The method for identifying the judgement area in the court trial video according to claim 1, characterized in that: After performing second feature recognition on the human-occupied area to obtain the platform sub-area, the method further includes: Select one of the French platform sub-areas as the lateral reference area; According to a preset step length, horizontally translate the boundary box of the horizontal reference area in the trial video screen to obtain a suspected judge platform sub-area; The second feature recognition is performed on the suspected law platform sub-region, and the suspected law platform sub-region in which the second feature is recognized is used as the law platform sub-region.

4. The method for identifying the judgement area in the court trial video according to claim 3, characterized in that: The lateral reference area is adjacent to only one bench sub-area in the horizontal direction.

5. The method for identifying the judgement area in the court trial video according to claim 3, characterized in that: Before determining the lateral reference area in the base sub-area, the method further includes: The judge platform sub-area close to the bottom edge of the trial video screen is used as the vertical reference area; Calculating the longitudinal distance between the bench sub-area excluding the longitudinal reference area and the longitudinal reference area; The platform sub-area whose longitudinal distance exceeds a second preset threshold value is removed.

6. The method for identifying the judgement area in the court trial video according to claim 1, characterized in that: The second feature recognition of the human-occupied area further includes: If the second feature is not recognized in the human-inhabited area, expanding the human-inhabited area according to a predetermined range; A second feature recognition is performed on the enlarged human-containing area.

7. The method for identifying the judgement area in the court trial video according to claim 1, characterized in that: Combining the judge platform sub-areas to obtain the judge platform area in the trial video screen further includes: Filtering out the boundary closest to the boundary of the trial video screen from the boundary of the judge platform sub-area; The area enclosed by the extended line of the screened boundary is used as the legal platform area.

8. A device for identifying the judge's platform area in a court trial video, characterized in that: include, A human region determination unit is used to perform first feature recognition on the trial video screen to obtain a human region, wherein the first feature is a human body feature; A judge's bench sub-area determination unit, configured to perform second feature recognition on the manned area to obtain a judge's bench sub-area, wherein the second feature is a predetermined feature of the judge; The judge table area determination unit is used to combine the judge table sub-areas to obtain the judge table area in the trial video screen, and the combination method is to splice according to the position relationship of each judge table sub-area in the trial video screen.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory, characterized in that: When the computer program is executed by the processor, the computer program executes the instructions of the method according to any one of claims 1 to 7.

10. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor of a computer device, the computer program executes instructions of the method according to any one of claims 1 to 7.

Citation Information

Patent Citations

  • Judge dressing detection method based on video

    CN105488490A