Video analysis intelligent AI management platform for industry

Abstract

The invention belongs to the technical field of video analysis, and particularly relates to a video analysis intelligent AI management platform for the industrial industry, which comprises a real-time monitoring module used for displaying pictures of each camera in real time and supporting multi-picture switching; the optimization module is used for optimizing the picture shot by the real-time monitoring module so as to improve the shooting definition of the real-time monitoring module; and the intelligent analysis module is used for analyzing personnel behaviors, personnel identities and equipment states. According to the invention, by arranging the personnel analysis module and the identity recognition module, when the personnel analysis module analyzes that the personnel have violation operation, the personnel can be subjected to face recognition, so that the personnel identity can be quickly known, and in addition, if the face recognition fails, the personnel can be tracked and recognized through the features, so that the operation efficiency is improved. And personnel are prevented from escaping and punishing.

Description

Technical Field

[0001] The present invention relates to the technical field of video analysis, and specifically provides an intelligent AI management platform for video analysis in the industrial industry. Background Art

[0002] The industrial industry refers to the social material production sectors that extract natural resources and process various raw materials. It is the main component of the secondary industry and has gone through development stages such as handicrafts and machine industry, and is divided into two categories: light industry and heavy industry. Among them, the intelligent AI management platform for video analysis in the industrial industry is a system that uses artificial intelligence technology to analyze and manage video data in the industrial production process, and can achieve real-time monitoring, intelligent early warning, data analysis and decision support of the industrial production process, improving the safety, efficiency and quality of industrial production.

[0003] Currently, in the existing intelligent AI management platform for video analysis in the industrial industry, the behavior of personnel is usually analyzed to determine whether there are any violations, such as not wearing a safety helmet. However, during this process, the identity of the violating personnel cannot be identified. Although there are systems in the current market that can identify the faces of personnel, this system cannot track the information of personnel when the face cannot be recognized, which has limitations. Therefore, an intelligent AI management platform for video analysis in the industrial industry is invented. Summary of the Invention

[0004] To solve the above technical problems, according to one aspect of the present invention, the following technical solutions are provided:

[0005] An intelligent AI management platform for video analysis in the industrial industry, which includes:

[0006] A real-time monitoring module for displaying the images of each camera in real time and supporting multi-screen switching;

[0007] An optimization module for optimizing the images captured by the real-time monitoring module to improve the clarity of the images captured by the real-time monitoring module;

[0008] An intelligent analysis module for analyzing personnel behavior, personnel identity and equipment status;

[0009] An early warning and alarm module for timely notifying relevant personnel by means of sound and light alarm, SMS, and email when the intelligent analysis module analyzes an abnormal situation;

[0010] The intelligent analysis module includes:

[0011] A personnel analysis module for analyzing personnel behavior based on the images captured by the real-time monitoring module to determine whether there are any violations.

[0012] An identity recognition module, which is used to recognize the identity of a person when it is analyzed that the person has violated the operation rules;

[0013] A device status monitoring module, which is used to analyze the operating device according to the images captured by the real-time monitoring module to analyze whether there are problems with the device;

[0014] The identity recognition module includes:

[0015] A face image acquisition module, which is used to acquire the face image of a person according to the images captured by the real-time monitoring module;

[0016] A face storage module, which is used to store the face images of each staff member;

[0017] A face comparison module, which is used to compare the images acquired by the face image acquisition module with the images stored in the face storage module;

[0018] A face judgment module, which is used to judge whether there is a similarity between the images acquired by the face image acquisition module and those stored in the face storage module. If there is a similarity, it means the recognition is successful; if not, it means the recognition fails;

[0019] A feature extraction module, which is used to extract the features of a person when face recognition fails;

[0020] A feature storage module, which is used to store the features extracted by the feature extraction module;

[0021] A comparison module, which is used to compare the images captured by the real-time monitoring module with the feature images stored in the feature storage module. If there is a similarity, it will be uploaded to the warning and alarm module.

[0022] As a preferred solution of the video analysis intelligent AI management platform for the industrial industry described in the present invention, wherein: the optimization module includes:

[0023] An image dust removal module, which is used to remove the dust background in the images captured by the real-time monitoring module;

[0024] An image shadow removal module, which is used to remove the shadow background in the images captured by the real-time monitoring module;

[0025] An image illumination adjustment module, which is used to adjust the illumination in the images captured by the real-time monitoring module.

[0026] As a preferred solution of the video analysis intelligent AI management platform for the industrial industry described in the present invention, wherein: the personnel analysis module includes:

[0027] A personnel image acquisition module, which is used to acquire the personnel features in the images captured by the real-time monitoring module;

[0028] A behavior acquisition module, which is used to acquire the actions, clothing, and positions of personnel according to the personnel features acquired by the personnel image acquisition module.

[0029] As a preferred solution of an intelligent AI management platform for video analysis in the industrial industry according to the present invention, wherein: the personnel analysis module further includes:

[0030] A behavior storage module, which is used to store various illegal operations of personnel;

[0031] A behavior comparison module, which is used to compare the actions, clothing, and positions of personnel acquired by the behavior acquisition module with the illegal operations of personnel stored in the behavior storage module;

[0032] A behavior judgment module, which is used to judge whether there are similarities in the actions, clothing, and positions of personnel acquired by the behavior acquisition module in the personnel storage module. If so, it means that the personnel belong to illegal operations and will be uploaded to the early warning and alarm module.

[0033] As a preferred solution of an intelligent AI management platform for video analysis in the industrial industry according to the present invention, wherein: the equipment status monitoring module includes:

[0034] An equipment image acquisition module, which is used to acquire the equipment features in the images captured by the real-time monitoring module;

[0035] An equipment storage module, which is used to store various damaged images of equipment.

[0036] As a preferred solution of an intelligent AI management platform for video analysis in the industrial industry according to the present invention, wherein: the equipment status monitoring module further includes:

[0037] An equipment comparison module, which is used to compare the equipment features acquired by the equipment image acquisition module with the equipment images stored in the equipment storage module;

[0038] An equipment judgment module, which is used to judge whether the equipment features acquired by the equipment image acquisition module are the same in the equipment storage module. If so, it means that there is a problem with the equipment and will be uploaded to the early warning and alarm module.

[0039] As a preferred solution of an intelligent AI management platform for video analysis in the industrial industry according to the present invention, wherein: it further includes:

[0040] A data analysis module, which is used to statistically analyze historical video data and analysis results, generate reports and charts, and provide data support for enterprise decision-making.

[0041] As a preferred solution of an intelligent AI management platform for video analysis in the industrial industry described in the present invention, it further includes:

[0042] A video management module, which is used to implement functions such as video storage, retrieval, playback, and download, facilitating users to manage video data.

[0043] Compared with the prior art:

[0044] 1. By setting a personnel analysis module and an identity recognition module, when the personnel analysis module analyzes that a person has violated the operation, face recognition of the person can be achieved, and then the identity of the person can be quickly known. In addition, if the face recognition fails, the person can also be tracked and recognized through features to prevent the person from escaping punishment;

[0045] 2. By setting an optimization module, when the real-time monitoring module acquires an image, dust, shadows, and light in the image can be removed and adjusted, thereby ensuring the clarity of the image, and thus improving the analysis accuracy of the video to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic flowchart of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0047] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0048] The present invention provides an intelligent AI management platform for video analysis in the industrial industry. Please refer to Figure 1 , including: a real-time monitoring module, which is used to display the images of each camera in real time and support multi-screen switching; an optimization module, which is used to optimize the images captured by the real-time monitoring module to improve the clarity of the images captured by the real-time monitoring module; an intelligent analysis module, which is used to analyze personnel behavior, personnel identity, and equipment status; an early warning and alarm module, which is used to timely notify relevant personnel by means of sound and light alarm, SMS, and email when the intelligent analysis module analyzes an abnormal situation; a data analysis module, which is used to statistically analyze historical video data and analysis results, generate reports and charts, and provide data support for enterprise decision-making; a video management module, which is used to implement functions such as video storage, retrieval, playback, and download, facilitating users to manage video data;

[0049] The optimization module includes: an image dust removal module for removing the dust background in the images captured by the real-time monitoring module; an image shadow removal module for removing the shadow background in the images captured by the real-time monitoring module; and an image illumination adjustment module for adjusting the illumination in the images captured by the real-time monitoring module.

[0050] The steps for the image dust removal module are as follows:

[0051] S1, Spatial filtering:

[0052] S11, Median filtering: Replace the current pixel value with the median of the neighboring pixels;

[0053] S12, Gaussian filtering: Blur the image by weighted averaging with a Gaussian kernel to weaken fine dust;

[0054] S13, Bilateral filtering: Denoise while preserving edges, suitable for scenarios with low contrast between dust and background;

[0055] S2, Temporal filtering:

[0056] S21, Temporal median filtering: Take the median of multiple consecutive frames to eliminate randomly appearing dust;

[0057] S22, Optical flow method + Motion compensation: Estimate the motion vector through optical flow and average the noise reduction for the same area of adjacent frames;

[0058] S3, Frequency domain processing:

[0059] S31, Fourier transform: Convert the image to the frequency domain and filter out high-frequency noise.

[0060] The steps for the image shadow removal module are as follows:

[0061] S1, Illumination compensation and enhancement:

[0062] S11, Homomorphic filtering: Separate the illumination and reflectance of the image and suppress the low-frequency illumination component;

[0063] S12, Retinex theory: Estimate the illumination component through multi-scale Gaussian blur and restore the reflectance image;

[0064] S2, Shadow detection and segmentation:

[0065] S21, Threshold-based method: Segment the shadow using the luminance channel in the HSV / Lab color space;

[0066] S22, Background difference method: Assume that the shadow area is darker than the background and detect the shadow through inter-frame difference;

[0067] S3, Restoration and interpolation:

[0068] S31, Image inpainting algorithms (such as Telea algorithm, Criminisi algorithm): Use the texture of the non-shadow area to fill the shadow.

[0069] The steps of the image illumination adjustment module are as follows:

[0070] S1, Regression model:

[0071] S11: Train a regression model (such as random forest) to predict illumination parameters (brightness, contrast, color temperature). The input features include: global statistics (mean, variance, histogram), local features (gradient, edge intensity);

[0072] S2, Clustering and classification:

[0073] S21, Use K-means to divide the image into high-light, middle-tone, and shadow regions, and adjust the parameters respectively;

[0074] S22, A classification model based on CNN (such as ResNet) is used to identify the illumination scene and match the preset adjustment strategy.

[0075] The intelligent analysis module includes: a personnel analysis module, which is used to analyze the behavior of personnel according to the images captured by the real-time monitoring module to analyze whether there are any violations by the personnel; an identity recognition module, which is used to identify the identity of the personnel when it is analyzed that there are violations by the personnel; a device status monitoring module, which is used to analyze the operating devices according to the images captured by the real-time monitoring module to analyze whether there are any problems with the devices;

[0076] The identity recognition module includes: a face image acquisition module, which is used to acquire the face images of personnel according to the images captured by the real-time monitoring module; a face storage module, which is used to store the face images of each staff member; a face comparison module, which is used to compare the images acquired by the face image acquisition module with the images stored in the face storage module; a face judgment module, which is used to judge whether there are similarities between the images acquired by the face image acquisition module in the face storage module. If there are similarities, it means that the recognition is successful. If there are no similarities, it means that the recognition fails; a feature extraction module, which is used to extract the features of the personnel when the face recognition fails; a feature storage module, which is used to store the features extracted by the feature extraction module; a comparison module, which is used to compare the images captured by the real-time monitoring module with the feature images stored in the feature storage module. If there are similarities, it will be uploaded to the warning and alarm module.

[0077] The personnel analysis module includes: a personnel image acquisition module for acquiring the personnel features in the images captured by the real-time monitoring module; a behavior acquisition module for acquiring the actions, clothing, and positions of personnel based on the personnel features acquired by the personnel image acquisition module; a behavior storage module for storing various illegal operations of personnel; a behavior comparison module for comparing the actions, clothing, and positions of personnel acquired by the behavior acquisition module with the illegal operations of personnel stored in the behavior storage module; and a behavior judgment module for judging whether there are similarities in the actions, clothing, and positions of personnel acquired by the behavior acquisition module in the personnel storage module. If so, it indicates that the personnel are engaged in illegal operations and will be uploaded to the warning and alarm module.

[0078] The equipment status monitoring module includes: an equipment image acquisition module for acquiring the equipment features in the images captured by the real-time monitoring module; an equipment storage module for storing various damaged images of the equipment; an equipment comparison module for comparing the equipment features acquired by the equipment image acquisition module with the equipment images stored in the equipment storage module; and an equipment judgment module for judging whether the equipment features acquired by the equipment image acquisition module are the same as those stored in the equipment storage module. If so, it indicates that there is a problem with the equipment and will be uploaded to the warning and alarm module.

[0079] In specific use, the operation steps of those skilled in the art are as follows:

[0080] Step 1: The real-time monitoring module is used to display the images of each camera in real time, supporting multi-screen switching, so as to monitor the industrial environment.

[0081] Step 2: When the image contains dust, the dust background in the image captured by the real-time monitoring module will be removed through the image dust removal module; when the image contains shadows, the shadow background in the image captured by the real-time monitoring module will be removed through the image shadow removal module; during this process, the illumination of the image captured by the real-time monitoring module will also be adjusted through the image illumination adjustment module.

[0082] Step 3: The personnel image acquisition module is used to acquire the personnel features in the images captured by the real-time monitoring module. After acquisition, the behavior acquisition module will acquire the actions, clothing, and positions of personnel based on the personnel features acquired by the personnel image acquisition module. Then, the behavior comparison module will compare the actions, clothing, and positions of personnel acquired by the behavior acquisition module with the illegal operations of personnel stored in the behavior storage module. After comparison, the behavior judgment module will judge whether there are similarities in the actions, clothing, and positions of personnel acquired by the behavior acquisition module in the personnel storage module. If so, it indicates that the personnel are engaged in illegal operations and will be uploaded to the warning and alarm module.

[0083] Step 4: The face image acquisition module acquires the face image of the person according to the image captured by the real-time monitoring module. After acquisition, the face comparison module compares the image acquired by the face image acquisition module with the image stored in the face storage module. After comparison, the face judgment module determines whether there is a similarity between the image acquired by the face image acquisition module and the image stored in the face storage module. If there is a similarity, it indicates successful recognition. If there is no similarity, it indicates failed recognition. If the recognition fails, the feature extraction module extracts the features of the person (such as height, body posture, shoe color, etc.). After extraction, the comparison module compares the image captured by the real-time monitoring module with the feature images stored in the feature storage module. If there is a similarity, it will be uploaded to the warning and alarm module;

[0084] Step 5: The device image acquisition module acquires the device features in the image captured by the real-time monitoring module. After acquisition, the device comparison module compares the device features acquired by the device image acquisition module with the device images stored in the device storage module. After comparison, the device judgment module determines whether the device features acquired by the device image acquisition module are the same as those stored in the device storage module. If they are the same, it indicates that there is a problem with the device and it will be uploaded to the warning and alarm module;

[0085] Step 6: When the intelligent analysis module analyzes an abnormal situation, the warning and alarm module can timely notify relevant personnel through sound and light alarms, text messages, and emails. After that, the data analysis module statistically analyzes the historical video data and analysis results, generates reports and charts, providing data support for enterprise decision-making. At the same time, the video management module also realizes functions such as video storage, retrieval, playback, and download, facilitating users to manage video data.

[0086] Although the present invention has been described above with reference to the embodiments, various improvements can be made to it and components therein can be replaced with equivalents without departing from the scope of the present invention. In particular, as long as there is no structural conflict, the features in the embodiments disclosed in the present invention can be combined with each other in any way. The exhaustive description of these combinations is not given in this specification only for the sake of saving space and resources. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. An intelligent AI management platform for video analysis in the industrial sector, characterized in that, Including: A real-time monitoring module, which is used to display the images of each camera in real time and supports multi-screen switching; An optimization module, which is used to optimize the images captured by the real-time monitoring module to improve the clarity of the images captured by the real-time monitoring module; An intelligent analysis module, which is used to analyze personnel behavior, personnel identity and equipment status; An early warning and alarm module, which is used to timely notify relevant personnel by means of sound and light alarm, SMS and email when the intelligent analysis module analyzes abnormal situations; The intelligent analysis module includes: A personnel analysis module, which is used to analyze personnel behavior according to the images captured by the real-time monitoring module to analyze whether there are any violations by personnel; An identity recognition module, which is used to recognize the identity of personnel when it is analyzed that there are violations by personnel; A device status monitoring module, which is used to analyze the running devices according to the images captured by the real-time monitoring module to analyze whether there are any problems with the devices; The identity recognition module includes: A face image acquisition module, which is used to acquire the face images of personnel according to the images captured by the real-time monitoring module; A face storage module, which is used to store the face images of each staff member; A face comparison module, which is used to compare the images acquired by the face image acquisition module with the images stored in the face storage module; A face judgment module, which is used to judge whether there are similarities between the images acquired by the face image acquisition module and those stored in the face storage module. If there are similarities, it means the recognition is successful. If not, it means the recognition fails; A feature extraction module, which is used to extract the features of personnel when face recognition fails; A feature storage module, which is used to store the features extracted by the feature extraction module; A comparison module, which is used to compare the images captured by the real-time monitoring module with the feature images stored in the feature storage module. If there are similarities, they will be uploaded to the early warning and alarm module; The optimization module includes: An image dust removal module, which is used to remove the dust background in the images captured by the real-time monitoring module; An image shadow removal module, which is used to remove the shadow background in the images captured by the real-time monitoring module; An image illumination adjustment module, which is used to adjust the illumination in the images captured by the real-time monitoring module.

2. The intelligent AI management platform for video analysis in the industrial sector according to claim 1, characterized in that The personnel analysis module includes: A personnel image acquisition module, which is used to acquire the personnel features in the images captured by the real-time monitoring module; A behavior acquisition module, which is used to acquire the actions, clothing and positions of personnel according to the personnel features acquired by the personnel image acquisition module.

3. An intelligent AI management platform for video analysis in the industrial sector according to claim 2, characterized in that, The personnel analysis module also includes: A behavior storage module, which is used to store various violations of personnel; A behavior comparison module, which is used to compare the actions, clothing and positions of personnel acquired by the behavior acquisition module with the violations of personnel stored in the behavior storage module; A behavior judgment module, which is used to judge whether the actions, clothing and positions of personnel acquired by the behavior acquisition module are similar to those stored in the personnel storage module. If there are similarities, it means the personnel belong to violations and will be uploaded to the early warning and alarm module.

4. An intelligent AI management platform for video analysis in the industrial sector according to claim 1, characterized in that, The device status monitoring module includes: The device image acquisition module is used to acquire the device features in the images captured by the real-time monitoring module; The device storage module is used to store various damaged images of the device.

5. The intelligent AI management platform for video analysis in the industrial sector according to claim 4, characterized in that, The device status monitoring module further includes: The device comparison module is used to compare the device features acquired by the device image acquisition module with the device images stored in the device storage module; The device judgment module is used to judge whether there are the same device features acquired by the device image acquisition module in the device storage module. If so, it indicates that there is a problem with the device and will be uploaded to the warning and alarm module.

6. The intelligent AI management platform for video analysis in the industrial sector according to claim 1, characterized in that, It further includes: The data analysis module is used to statistically analyze the historical video data and analysis results, generate reports and charts, and provide data support for enterprise decision-making.

7. An intelligent AI management platform for video analysis in the industrial sector according to claim 1, characterized in that, It further includes: The video management module is used to implement functions such as video storage, retrieval, playback, and download, facilitating users to manage video data.

Images