A method and system for intelligent information quality control based on quality perception
Through the intelligent control method of information quality based on quality perception, multi-level quality evaluation and regional expansion are used for binocular cameras, the problem of regional missed detection in 3D data detection is solved, and detailed evaluation and timely alarm of image quality are achieved to ensure that The security of the equipment.
Patent Information
- Application Number
- CN202411224750.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-03
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-09-03
AI Technical Summary
The prior art can easily cause regional missed detection in 3D data quality detection, resulting in data pileup during image repair and fail to effectively process areas with poor image quality.
Using the intelligent information quality control method based on quality perception, the target image information is obtained through a binocular camera, the evaluation area collection is constructed, multi-level quality evaluation and area expansion are carried out, the low-quality areas in the image are marked, and the visual prompt is sent based on the evaluation results. or alarm.
It realizes a more complete quality evaluation of 3D image data, improves the efficiency of calculation and analysis, promptly discovers and prompts image quality problems, prevents equipment damage, and ensures asset safety.
Smart Images

Figure CN118735920B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of intelligent information quality control, and specifically to an intelligent information quality control method and system based on quality perception. Background Art
[0002] With the development of the times, the dissemination form of content is constantly evolving, and upgrading from 2D to 3D is the inevitable direction of technological development. 3D imaging technology has penetrated the market comprehensively. Binoculars simulate human eyes. When observing the same object, the two eyes see different objects. Through this difference, 3D data can be calculated.
[0003] In the quality detection of 3D data, the same evaluation method is often used to evaluate the quality of the entire image data, which can easily cause regional missed detections and affect subsequent 3D reconstruction. The existing technology does not uniformly process areas with poor image quality, which can easily cause data stacking during subsequent image restoration.
[0004] Therefore, the present invention discloses an information quality intelligent management and control method and system based on quality perception to solve the above problems. Summary of the invention
[0005] The purpose of the present invention is to provide a method and system for intelligent information quality control based on quality perception to solve the problems raised in the above background technology.
[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for intelligent information quality control based on quality perception, characterized in that the method comprises the following steps:
[0007] S1: photographing a target object with a binocular camera to obtain target image information, constructing a target image information data set, obtaining a target image pixel matrix based on the target image information, associating the target image pixel matrices of the left camera and the right camera at the same time, generating a target image pixel matrix set, and generating a disparity image pixel matrix based on the target image pixel matrix;
[0008] S2: constructing an evaluation region set of target image information, performing a first region quality evaluation on the disparity image pixel matrix according to the evaluation region set, extracting a corresponding part of the pixel matrix in the target image pixel matrix based on the first region quality evaluation value, and performing a second region quality evaluation;
[0009] S3: Expand the area of the elements in the evaluation area set of the target image information, perform a third area quality evaluation on the corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation area, compare the second area quality evaluation with the third area quality evaluation, and mark in the target image pixel matrix according to the comparison result;
[0010] S4: Evaluate the image information quality based on the mark, and send a visual prompt or alarm to the user according to the image information quality evaluation result.
[0011] According to the above scheme, in step S1, each target image information in the target image information data set is uniformly labeled, and the target image information data set is recorded as TII={TII i |i∈[1,n]}, where TII i represents the target image information with sequence number i in the target image information data set, and n represents the total number of target image information;
[0012] The associated image pixel matrix set is denoted as TIP={TIP i |i∈[1,n]}; where the target image pixel matrix TIP is numbered i i Based on target image information TII i Get, the target image pixel matrix includes the left camera target image pixel matrix and the right camera target image pixel matrix; the left camera target image pixel matrix is recorded as TIP i L , the pixel value of the pixel at the xth row and yth column in the left camera target image pixel matrix is recorded as TIP i L (x,y) ; The right camera target image pixel matrix is recorded as TIP i R , the pixel value of the pixel at the xth row and yth column in the pixel matrix of the right camera target image is recorded as TIP i R (x,y) ; Associate the left camera target image pixel matrix and the right camera target image pixel matrix at the same time so that the pixel coordinates of the two target image pixel matrices correspond one to one;
[0013] Based on the left camera target image pixel matrix TIP in the target image pixel matrix i L and the right camera target image pixel matrix TIP i R Generate disparity image pixel matrix PIP i .
[0014] According to the above scheme, in step S2, the evaluation area set is recorded as EA={EA a |a∈[1,A]}, where EA a The evaluation area with serial number a in the evaluation area set, A is the total number of evaluation areas;
[0015] The evaluation area includes a sight discontinuity area, a non-blocked area, an edge area, a background area and a foreground area; a target image pixel matrix may contain more than one evaluation area of the same kind;
[0016] Through different types of evaluation areas, regional quality evaluation can be performed on different areas in the target image pixel matrix, and regional quality evaluation can be performed on the target image pixel matrix more completely;
[0017] Because there is no overlap between the various evaluation areas, the same target image pixel matrix can be detected at the same time. According to different evaluation areas, multi-process technology is used to allow the calculation and analysis tasks of multiple evaluation areas to be performed simultaneously, which can effectively improve the efficiency of calculation and analysis.
[0018] According to the evaluation area EA a , the disparity image pixel matrix is evaluated for the first region quality, and the specific calculation formula is as follows:
[0019] ;
[0020] PQE i a Indicates the evaluation area EA a In the disparity image pixel matrix PIP i The first area quality evaluation value, N EAa Indicates the evaluation area EA a The number of pixels in PIP i(x,y) Represents the disparity image pixel matrix PIP i The pixel value of the pixel at the xth row and yth column in ;
[0021] The quality evaluation value PQE of the first area of the image i a Set the threshold value. The threshold value of the first area quality evaluation value of different evaluation areas is different. If the image first area quality evaluation value PQE i a is greater than or equal to the set threshold, extract the evaluation area pixel matrix in the disparity image pixel matrix; if the image first area quality evaluation value PQE i a If it is less than the set threshold, no processing will be done;
[0022] Although there is a certain position deviation between the left camera target image and the right camera target image, in theory, the pixel difference between them is not large. If there is a region with a high pixel value in the parallax image between the left camera target image and the right camera target image, it means that there is interference and loss in the shooting or transmission process, which reduces the image quality.
[0023] According to the above method, the quality evaluation value of the first region is calculated and the threshold is set, so that the region with higher pixel values in the disparity image pixel matrix, that is, the region with poor image quality, can be effectively extracted;
[0024] In the original left camera target image pixel matrix and the right camera target image pixel matrix of the disparity image pixel matrix, the evaluation area is mapped to the corresponding left camera target image pixel matrix and the right camera target image pixel matrix; and the projected left camera evaluation area pixel matrix and the right camera evaluation area pixel matrix are extracted;
[0025] The second area quality evaluation is performed on the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area. The specific calculation formula is as follows:
[0026] ;
[0027] TLQE i a Represents the second area quality evaluation value of the evaluation area in the left camera evaluation area pixel matrix,
[0028] ;
[0029] TRQE i a Represents the second area quality evaluation value of the evaluation area in the pixel matrix of the right camera evaluation area.
[0030] According to the above scheme, in step S3, the region EA is evaluated a The geometric center of is the midpoint, and the area EA will be evaluated a Expand out the preset expansion ratio of the system, and record the expanded evaluation area as EEA a , based on the expanded evaluation area, the third area quality evaluation is performed on the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area. The specific calculation formula is as follows:
[0031] ;
[0032] ;
[0033] TTLQE i a Indicates the third area quality evaluation value of the evaluation area in the left camera evaluation area pixel matrix; TTRQE i a Represents the third area quality evaluation value of the evaluation area in the pixel matrix of the right camera evaluation area;
[0034] The proportional coefficient between the second area quality evaluation value and the third area quality evaluation value is calculated for the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area respectively. The specific calculation formula is as follows:
[0035] S L a =TLQE i a / TTLQE i a ;
[0036] S R a =TRQE i a / TTRQE i a ;
[0037] Where S L a represents the scale factor of the pixel matrix of the left camera evaluation area, S R a Represents the scale factor of the pixel matrix of the right camera evaluation area;
[0038] Set a threshold for the proportional coefficient. If S L a or S R a If any of the proportional coefficients is greater than or equal to the threshold, the evaluation area before this expansion is marked in the corresponding target image pixel matrix; if there is no S L a and S R a If both are less than the threshold, no processing will be performed;
[0039] Repeat step S3 until S L a and S R a are all less than the threshold.
[0040] Expand the evaluation area outward by the preset expansion ratio of the system, and repeat step S3 until S L a and S R a All of them are smaller than the threshold, which can find the areas with low image quality to the greatest extent and provide more complete basic information for subsequent repair;
[0041] According to the above scheme, in step S4, based on other evaluation areas, the operations of step S2 and step S3 are performed to mark all evaluation areas that meet the conditions, and the evaluation areas marked last constitute a marked evaluation area set, which is recorded as MEA={MEA b|b∈[1,B]}; where MEA b represents the marked evaluation area with serial number b, B represents the total number of evaluation area sets, B≤A;
[0042] The image information quality is evaluated based on the set of marked evaluation areas. The specific calculation formula is as follows:
[0043] ;
[0044] Where α is the image information quality evaluation value, Indicates the evaluation area MEA b The number of pixels in Represents the number of pixels in the left camera target image pixel matrix, Represents the number of pixels in the pixel matrix of the right camera target image;
[0045] A threshold is set for the image information quality evaluation. If the image information quality evaluation value is greater than or equal to the threshold, a visual prompt and an alarm are sent to the user; if the image information quality evaluation value is less than the threshold, only a visual prompt is sent to the user; the visual prompt includes a marked target image pixel matrix.
[0046] If the image information quality evaluation value is greater than or equal to the threshold, it means that the environment is no longer affecting the camera's posture, but there may be problems with the camera itself or the transmission line. It is necessary to send visual prompts and alarms to the user to prompt the user to inspect the shooting equipment to prevent further damage to the equipment and ensure the safety of the user's assets.
[0047] Another aspect of the present application provides an intelligent information quality control system based on quality perception, which is applied to the above-mentioned intelligent information quality control method based on quality perception, and is characterized in that the system includes a data collection and sorting module, a regional quality evaluation module, an evaluation extension marking module and an information quality evaluation prompt module;
[0048] The data acquisition and sorting module is used to shoot the target object with a binocular camera, obtain target image information, construct a target image information data set, obtain a target image pixel matrix based on the target image information, associate the target image pixel matrices of the left camera and the right camera at the same time, generate a target image pixel matrix set, and generate a parallax image pixel matrix based on the target image pixel matrix;
[0049] The regional quality evaluation module is used to construct an evaluation region set of target image information, perform a first regional quality evaluation on the disparity image pixel matrix according to the evaluation region set, extract a corresponding part of the pixel matrix in the target image pixel matrix based on the first regional quality evaluation value, and perform a second regional quality evaluation;
[0050] The evaluation extension marking module is used to perform regional expansion on the elements in the evaluation region set of the target image information, perform a third regional quality evaluation on the corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation region, compare the second regional quality evaluation with the third regional quality evaluation, and mark in the target image pixel matrix according to the comparison result;
[0051] The information quality evaluation prompt module is used to evaluate the image information quality based on the mark, and send a visual prompt or alarm to the user according to the image information quality evaluation result.
[0052] According to the above solution, the data acquisition and sorting module includes a data acquisition unit and a data sorting unit;
[0053] The data acquisition unit is used to photograph the target object using a binocular camera, obtain target image information, and construct a target image information data set;
[0054] The data sorting unit obtains a target image pixel matrix based on the target image information, associates the target image pixel matrices of the left camera and the right camera at the same time, generates a target image pixel matrix set, and generates a parallax image pixel matrix based on the target image pixel matrix.
[0055] According to the above solution, the regional quality assessment module includes a first regional quality assessment unit and a second regional quality assessment unit;
[0056] The first region quality evaluation unit is used to construct an evaluation region set of target image information, and perform a first region quality evaluation on the disparity image pixel matrix according to the evaluation region set;
[0057] The second region quality evaluation unit extracts a corresponding part of the pixel matrix in the target image pixel matrix based on the first region quality evaluation value, and performs a second region quality evaluation.
[0058] According to the above scheme, the evaluation extension marking module includes an evaluation area expansion unit and an evaluation comparison marking unit;
[0059] The evaluation region expansion unit is used to expand the region of the elements in the evaluation region set of the target image information, and perform a third region quality evaluation on a corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation region;
[0060] The evaluation comparison and marking unit is used to compare the second region quality evaluation and the third region quality evaluation, and mark in the target image pixel matrix according to the comparison result.
[0061] According to the above scheme, the information quality evaluation prompt module includes an information quality evaluation unit and a prompt alarm unit;
[0062] The information quality evaluation unit is used to perform image information quality evaluation numerical calculation based on the mark;
[0063] The prompt alarm unit is used to set a threshold for the image information quality evaluation. If the image information quality evaluation value is greater than or equal to the threshold, a visual prompt and an alarm are sent to the user; if the image information quality evaluation value is less than the threshold, only a visual prompt is sent to the user; the visual prompt includes a marked target image pixel matrix.
[0064] Compared with the prior art, the beneficial effects achieved by the present invention are: through different types of evaluation areas, regional quality evaluation can be performed on different areas in the target image pixel matrix, and the regional quality evaluation of the target image pixel matrix can be performed more completely; according to different evaluation areas, multi-process technology is used to allow the calculation and analysis tasks of multiple evaluation areas to be performed simultaneously, which can effectively improve the efficiency of calculation and analysis; the quality evaluation value of the first area is calculated, and a threshold is set, which can effectively extract the area with higher pixel values in the disparity image pixel matrix, that is, the area with poor image quality; for the image information quality evaluation value greater than or equal to the threshold, it means that it is no longer the environment that affects the shooting due to the posture of the camera, but there may be problems with the camera itself or the transmission line, and it is necessary to send visual prompts and alarms to the user to prompt the user to repair the shooting equipment to prevent further damage to the equipment, thereby ensuring the safety of the user's assets. BRIEF DESCRIPTION OF THE DRAWINGS
[0065] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:
[0066] Figure 1 It is a flow chart of an intelligent information quality control method and system based on quality perception of the present invention;
[0067] Figure 2 It is a structural schematic diagram of an information quality intelligent control method and system based on quality perception of the present invention. DETAILED DESCRIPTION
[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0069] See also Figure 1 The present invention provides a technical solution: an intelligent information quality control method based on quality perception, characterized in that the method comprises the following steps:
[0070] S1: photographing a target object with a binocular camera to obtain target image information, constructing a target image information data set, obtaining a target image pixel matrix based on the target image information, associating the target image pixel matrices of the left camera and the right camera at the same time, generating a target image pixel matrix set, and generating a disparity image pixel matrix based on the target image pixel matrix;
[0071] In step S1, each target image information in the target image information data set is uniformly labeled, and the target image information data set is recorded as TII={TII i |i∈[1,n]}, where TII i represents the target image information with sequence number i in the target image information data set, and n represents the total number of target image information;
[0072] The associated image pixel matrix set is denoted as TIP={TIP i |i∈[1,n]}; where the target image pixel matrix TIP is numbered i i Based on target image information TII i Get, the target image pixel matrix includes the left camera target image pixel matrix and the right camera target image pixel matrix; the left camera target image pixel matrix is recorded as TIP i L , the pixel value of the pixel at the xth row and yth column in the left camera target image pixel matrix is recorded as TIP i L (x,y) ; The right camera target image pixel matrix is recorded as TIP i R , the pixel value of the pixel at the xth row and yth column in the pixel matrix of the right camera target image is recorded as TIP i R (x,y) ; Associate the left camera target image pixel matrix and the right camera target image pixel matrix at the same time so that the pixel coordinates of the two target image pixel matrices correspond one to one;
[0073] Based on the left camera target image pixel matrix TIP in the target image pixel matrix i L and the right camera target image pixel matrix TIP i R Generate disparity image pixel matrix PIP i .
[0074] S2: constructing an evaluation region set of target image information, performing a first region quality evaluation on the disparity image pixel matrix according to the evaluation region set, extracting a corresponding part of the pixel matrix in the target image pixel matrix based on the first region quality evaluation value, and performing a second region quality evaluation;
[0075] In step S2, the evaluation area set is recorded as EA={EA a |a∈[1,A]}, where EA a The evaluation area with serial number a in the evaluation area set, A is the total number of evaluation areas;
[0076] According to the evaluation area EA a , the disparity image pixel matrix is evaluated for the first region quality, and the specific calculation formula is as follows:
[0077] ;
[0078] PQE i a Indicates the evaluation area EA a In the disparity image pixel matrix PIP i The first area quality evaluation value, N EAa Indicates the evaluation area EA a The number of pixels in PIP i(x,y) Represents the disparity image pixel matrix PIP i The pixel value of the pixel at the xth row and yth column in ;
[0079] The quality evaluation value PQE of the first area of the image i a Set the threshold value. The threshold value of the first area quality evaluation value of different evaluation areas is different. If the image first area quality evaluation value PQE i a is greater than or equal to the set threshold, extract the evaluation area pixel matrix in the disparity image pixel matrix; if the image first area quality evaluation value PQE i a If it is less than the set threshold, no processing will be done;
[0080] In the original left camera target image pixel matrix and the right camera target image pixel matrix of the disparity image pixel matrix, the evaluation area is mapped to the corresponding left camera target image pixel matrix and the right camera target image pixel matrix; and the projected left camera evaluation area pixel matrix and the right camera evaluation area pixel matrix are extracted;
[0081] The second area quality evaluation is performed on the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area. The specific calculation formula is as follows:
[0082] ;
[0083] TLQE i a Represents the second area quality evaluation value of the evaluation area in the left camera evaluation area pixel matrix,
[0084] ;
[0085] TRQE i a Represents the second area quality evaluation value of the evaluation area in the pixel matrix of the right camera evaluation area.
[0086] S3: Expand the area of the elements in the evaluation area set of the target image information, perform a third area quality evaluation on the corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation area, compare the second area quality evaluation with the third area quality evaluation, and mark in the target image pixel matrix according to the comparison result;
[0087] In step S3, the evaluation area EA a The geometric center of is the midpoint, and the area EA will be evaluated a Expand out the preset expansion ratio of the system, and record the expanded evaluation area as EEA a , based on the expanded evaluation area, the third area quality evaluation is performed on the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area. The specific calculation formula is as follows:
[0088] ;
[0089] ;
[0090] TTLQE i a Indicates the third area quality evaluation value of the evaluation area in the left camera evaluation area pixel matrix; TTRQE i a Represents the third area quality evaluation value of the evaluation area in the pixel matrix of the right camera evaluation area;
[0091] The proportional coefficient between the second area quality evaluation value and the third area quality evaluation value is calculated for the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area respectively. The specific calculation formula is as follows:
[0092] S L a =TLQE i a / TTLQE i a ;
[0093] S R a =TRQE i a / TTRQE i a ;
[0094] Where S L a represents the scale factor of the pixel matrix of the left camera evaluation area, S R a Represents the scale factor of the pixel matrix of the right camera evaluation area;
[0095] Set a threshold for the proportional coefficient. If S L a or S R a If any of the proportional coefficients is greater than or equal to the threshold, the evaluation area before this expansion is marked in the corresponding target image pixel matrix; if there is no S L a and S R a If both are less than the threshold, no processing will be performed;
[0096] Repeat step S3 until S L a and S R a are all less than the threshold.
[0097] S4: Evaluate the image information quality based on the mark, and send a visual prompt or alarm to the user according to the image information quality evaluation result.
[0098] In step S4, based on other evaluation areas, the operations of step S2 and step S3 are performed to mark all evaluation areas that meet the conditions, and the evaluation areas marked last time are combined into a marked evaluation area set, which is recorded as MEA={MEA b |b∈[1,B]}; where MEA b represents the marked evaluation area with serial number b, B represents the total number of evaluation area sets, B≤A;
[0099] The image information quality is evaluated based on the set of marked evaluation areas. The specific calculation formula is as follows:
[0100] ;
[0101] Where α is the image information quality evaluation value, Indicates the evaluation area MEA b The number of pixels in Represents the number of pixels in the left camera target image pixel matrix, Represents the number of pixels in the pixel matrix of the right camera target image;
[0102] A threshold is set for the image information quality evaluation. If the image information quality evaluation value is greater than or equal to the threshold, a visual prompt and an alarm are sent to the user; if the image information quality evaluation value is less than the threshold, only a visual prompt is sent to the user; the visual prompt includes a marked target image pixel matrix.
[0103] See also Figure 2 ,The present invention provides a technical solution: an intelligent information quality management and control system based on quality perception, the system includes a data collection and sorting module, a regional quality evaluation module, an evaluation extension marking module and an information quality evaluation prompt module;
[0104] The data acquisition and sorting module is used to shoot the target object with a binocular camera, obtain target image information, construct a target image information data set, obtain a target image pixel matrix based on the target image information, associate the target image pixel matrices of the left camera and the right camera at the same time, generate a target image pixel matrix set, and generate a parallax image pixel matrix based on the target image pixel matrix;
[0105] The regional quality evaluation module is used to construct an evaluation region set of target image information, perform a first regional quality evaluation on the disparity image pixel matrix according to the evaluation region set, extract a corresponding part of the pixel matrix in the target image pixel matrix based on the first regional quality evaluation value, and perform a second regional quality evaluation;
[0106] The evaluation extension marking module is used to perform regional expansion on the elements in the evaluation region set of the target image information, perform a third regional quality evaluation on the corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation region, compare the second regional quality evaluation with the third regional quality evaluation, and mark in the target image pixel matrix according to the comparison result;
[0107] The information quality evaluation prompt module is used to evaluate the image information quality based on the mark, and send a visual prompt or alarm to the user according to the image information quality evaluation result.
[0108] According to the above solution, the data acquisition and sorting module includes a data acquisition unit and a data sorting unit;
[0109] The data acquisition unit is used to photograph the target object using a binocular camera, obtain target image information, and construct a target image information data set;
[0110] The data sorting unit obtains a target image pixel matrix based on the target image information, associates the target image pixel matrices of the left camera and the right camera at the same time, generates a target image pixel matrix set, and generates a parallax image pixel matrix based on the target image pixel matrix.
[0111] According to the above solution, the regional quality assessment module includes a first regional quality assessment unit and a second regional quality assessment unit;
[0112] The first region quality evaluation unit is used to construct an evaluation region set of target image information, and perform a first region quality evaluation on the disparity image pixel matrix according to the evaluation region set;
[0113] The second region quality evaluation unit extracts a corresponding part of the pixel matrix in the target image pixel matrix based on the first region quality evaluation value, and performs a second region quality evaluation.
[0114] According to the above scheme, the evaluation extension marking module includes an evaluation area expansion unit and an evaluation comparison marking unit;
[0115] The evaluation region expansion unit is used to expand the region of the elements in the evaluation region set of the target image information, and perform a third region quality evaluation on a corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation region;
[0116] The evaluation comparison and marking unit is used to compare the second region quality evaluation and the third region quality evaluation, and mark in the target image pixel matrix according to the comparison result.
[0117] According to the above scheme, the information quality evaluation prompt module includes an information quality evaluation unit and a prompt alarm unit;
[0118] The information quality evaluation unit is used to perform image information quality evaluation numerical calculation based on the mark;
[0119] The prompt alarm unit is used to set a threshold for the image information quality evaluation. If the image information quality evaluation value is greater than or equal to the threshold, a visual prompt and an alarm are sent to the user; if the image information quality evaluation value is less than the threshold, only a visual prompt is sent to the user; the visual prompt includes a marked target image pixel matrix.
[0120] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0121] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A quality-aware intelligent information quality control method, characterized in that: The method comprises the following steps: S1: photographing a target object with a binocular camera to obtain target image information, constructing a target image information data set, obtaining a target image pixel matrix based on the target image information, associating the target image pixel matrices of the left camera and the right camera at the same time, generating a target image pixel matrix set, and generating a disparity image pixel matrix based on the target image pixel matrix; S2: constructing an evaluation region set of target image information, performing a first region quality evaluation on the disparity image pixel matrix according to the evaluation region set, extracting a corresponding part of the pixel matrix in the target image pixel matrix based on the first region quality evaluation value, and performing a second region quality evaluation; S3: Expand the elements in the evaluation area set of the target image information, perform a third area quality evaluation on the corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation area, compare the second area quality evaluation with the third area quality evaluation, and mark in the target image pixel matrix according to the comparison result; S4: Evaluate the image information quality based on the mark, and send a visual prompt or alarm to the user according to the image information quality evaluation result; In the step S2, the evaluation area set is recorded as EA={EA a |a∈[1,A]}, where EA a The evaluation area with serial number a in the evaluation area set, A is the total number of evaluation areas; According to the evaluation area EA a , the disparity image pixel matrix is evaluated for the first region quality, and the specific calculation formula is as follows: ; PQE i a Indicates the evaluation area EA a In the disparity image pixel matrix PIP i The first area quality evaluation value, N EAa Indicates the evaluation area EA a The number of pixels in PIP i(x,y) Represents the disparity image pixel matrix PIP i The pixel value of the pixel at the xth row and yth column in ; The quality evaluation value PQE of the first area of the image i a Set the threshold value. The threshold value of the first area quality evaluation value of different evaluation areas is different. If the image first area quality evaluation value PQE i a is greater than or equal to the set threshold, extract the evaluation area pixel matrix in the disparity image pixel matrix; if the image first area quality evaluation value PQE i a If it is less than the set threshold, no processing will be done; In the original left camera target image pixel matrix and the right camera target image pixel matrix of the disparity image pixel matrix, the evaluation area is mapped to the corresponding left camera target image pixel matrix and the right camera target image pixel matrix; and the projected left camera evaluation area pixel matrix and the right camera evaluation area pixel matrix are extracted; The second area quality evaluation is performed on the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area. The specific calculation formula is as follows: ; TLQE i a Represents the second area quality evaluation value of the evaluation area in the left camera evaluation area pixel matrix, ; TRQE i a Represents the second area quality evaluation value of the evaluation area in the pixel matrix of the right camera evaluation area; In step S3, the evaluation area EA a The geometric center of is the midpoint, and the area EA will be evaluated a Expand out the preset expansion ratio of the system, and record the expanded evaluation area as EEA a , based on the expanded evaluation area, the third area quality evaluation is performed on the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area. The specific calculation formula is as follows: ; ; TTLQE i a Indicates the third area quality evaluation value of the evaluation area in the left camera evaluation area pixel matrix; TTRQE i a Represents the third area quality evaluation value of the evaluation area in the pixel matrix of the right camera evaluation area; The proportional coefficient between the second area quality evaluation value and the third area quality evaluation value is calculated for the pixel matrix of the left camera evaluation area and the pixel matrix of the right camera evaluation area respectively. The specific calculation formula is as follows: S L a =TLQE i a / TTLQE i a ; S R a =TRQE i a / TTRQE i a ; Where S L a represents the scale factor of the pixel matrix of the left camera evaluation area, S R a Represents the scale factor of the pixel matrix of the right camera evaluation area; Set a threshold for the proportional coefficient. If S L a or S R a If any of the proportional coefficients is greater than or equal to the threshold, the evaluation area before this expansion is marked in the corresponding target image pixel matrix; if there is no S L a and S R a If both are less than the threshold, no processing will be performed; Repeat step S3 until S L a and S R a All are less than the threshold; In step S4, based on other evaluation areas, the operations of step S2 and step S3 are performed to mark all evaluation areas that meet the conditions, and the evaluation areas marked last time are combined into a marked evaluation area set, which is recorded as MEA={MEA b |b∈[1,B]}; where MEA b represents the marked evaluation area with serial number b, B represents the total number of evaluation area sets, B≤A; The image information quality is evaluated based on the set of marked evaluation areas. The specific calculation formula is as follows: ; Where α is the image information quality evaluation value, N MEAb Indicates the evaluation area MEA b The number of pixels in L,i all Represents the number of pixels in the left camera target image pixel matrix, N R,i all Represents the number of pixels in the pixel matrix of the right camera target image; A threshold is set for the image information quality evaluation. If the image information quality evaluation value is greater than or equal to the threshold, a visual prompt and an alarm are sent to the user; If the image information quality evaluation value is less than the threshold, only a visual prompt is sent to the user; the visual prompt includes a marked target image pixel matrix.
2. According to claim 1, the method for intelligent information quality control based on quality perception is characterized by: In step S1, each target image information in the target image information data set is uniformly labeled, and the target image information data set is recorded as TII={TII i |i∈[1,n]}, where TII i represents the target image information with sequence number i in the target image information data set, and n represents the total number of target image information; The associated image pixel matrix set is denoted as TIP={TIP i |i∈[1,n]}; The target image pixel matrix TIP is numbered i. i Based on target image information TII i Get, the target image pixel matrix includes the left camera target image pixel matrix and the right camera target image pixel matrix; the left camera target image pixel matrix is recorded as TIP i L , the pixel value of the pixel at the xth row and yth column in the left camera target image pixel matrix is recorded as TIP i L (x,y) ; The right camera target image pixel matrix is recorded as TIP i R , the pixel value of the pixel at the xth row and yth column in the pixel matrix of the right camera target image is recorded as TIP i R (x,y) ; Associate the left camera target image pixel matrix and the right camera target image pixel matrix at the same time so that the pixel coordinates of the two target image pixel matrices correspond one to one; Based on the left camera target image pixel matrix TIP in the target image pixel matrix i L and the right camera target image pixel matrix TIP i R Generate disparity image pixel matrix PIP i .
3. An intelligent information quality control system based on quality perception, the system is applied to the intelligent information quality control method based on quality perception according to any one of claims 1-2, characterized in that: The system includes a data collection and sorting module, a regional quality evaluation module, an evaluation extension marking module and an information quality evaluation prompt module; The data acquisition and sorting module is used to shoot the target object with a binocular camera, obtain target image information, construct a target image information data set, obtain a target image pixel matrix based on the target image information, associate the target image pixel matrices of the left camera and the right camera at the same time, generate a target image pixel matrix set, and generate a parallax image pixel matrix based on the target image pixel matrix; The regional quality evaluation module is used to construct an evaluation region set of target image information, perform a first regional quality evaluation on the disparity image pixel matrix according to the evaluation region set, extract a corresponding part of the pixel matrix in the target image pixel matrix based on the first regional quality evaluation value, and perform a second regional quality evaluation; The evaluation extension marking module is used to perform regional expansion on the elements in the evaluation region set of the target image information, perform a third regional quality evaluation on the corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation region, compare the second regional quality evaluation with the third regional quality evaluation, and mark in the target image pixel matrix according to the comparison result; The information quality evaluation prompt module is used to evaluate the image information quality based on the mark, and send a visual prompt or alarm to the user according to the image information quality evaluation result.
4. According to claim 3, the information quality intelligent management and control system based on quality perception is characterized by: The data acquisition and sorting module includes a data acquisition unit and a data sorting unit; The data acquisition unit is used to photograph the target object using a binocular camera, obtain target image information, and construct a target image information data set; The data sorting unit obtains a target image pixel matrix based on the target image information, associates the target image pixel matrices of the left camera and the right camera at the same time, generates a target image pixel matrix set, and generates a parallax image pixel matrix based on the target image pixel matrix.
5. According to claim 3, the information quality intelligent management and control system based on quality perception is characterized by: The regional quality assessment module includes a first regional quality assessment unit and a second regional quality assessment unit; The first region quality evaluation unit is used to construct an evaluation region set of target image information, and perform a first region quality evaluation on the disparity image pixel matrix according to the evaluation region set; The second region quality evaluation unit extracts a corresponding part of the pixel matrix in the target image pixel matrix based on the first region quality evaluation value, and performs a second region quality evaluation.
6. The information quality intelligent management and control system based on quality perception according to claim 3 is characterized by: The evaluation extension marking module includes an evaluation area expansion unit and an evaluation comparison marking unit; The evaluation region expansion unit is used to expand the region of the elements in the evaluation region set of the target image information, and perform a third region quality evaluation on a corresponding part of the pixel matrix in the target image pixel matrix based on the expanded evaluation region; The evaluation comparison and marking unit is used to compare the second region quality evaluation and the third region quality evaluation, and mark in the target image pixel matrix according to the comparison result.
7. The information quality intelligent management and control system based on quality perception according to claim 3 is characterized by: The information quality evaluation prompt module includes an information quality evaluation unit and a prompt alarm unit; The information quality evaluation unit is used to perform image information quality evaluation numerical calculation based on the mark; The prompt alarm unit is used to set a threshold for the image information quality evaluation, and if the image information quality evaluation value is greater than or equal to the threshold, a visual prompt and alarm are sent to the user; If the image information quality evaluation value is less than the threshold, only a visual prompt is sent to the user; the visual prompt includes a marked target image pixel matrix.
