Computer image processing system based on artificial intelligence model

Abstract

The invention relates to a computer image processing system based on an artificial intelligence model. The computer image processing system comprises a time analysis device, a visual monitor, a sequence mapping device, a content identification mechanism, a region processing mechanism, a directional prediction mechanism, an optimization processing mechanism and a data playing mechanism. The computer image processing system based on the artificial intelligence model is compact in logic and intelligent in operation.

Description

Technical Field

[0001] The present invention relates to the field of computer processing, and in particular to a computer image processing system based on an artificial intelligence model. Background Art

[0002] Signal processing is the basic theory and technology of telecommunications. Its mathematical theories include equation theory, function theory, number theory, random process theory, least square method and optimization theory, and its technical pillars are circuit analysis, synthesis and electronic computer technology. Signal processing is closely related to contemporary pattern recognition, artificial intelligence, neural network computing and multimedia information processing, and it closely links basic theory with engineering applications. Therefore, signal processing is a discipline that has both a complex mathematical analysis background and broad practical engineering prospects.

[0003] Signal processing is developed with digital signal processing as the center. This is because signals can generally be represented in digital form, and digital signals can be calculated or processed on electronic computers through software. In this way, no matter how complex the calculation is, as long as it can be analyzed mathematically and the optimal solution can be obtained, it can be simulated on an electronic computer. If the calculation speed is appropriately fast, it can also be completed in real time using ultra-large-scale dedicated digital signal processing chips. Therefore, digital signal processing technology has become one of the most dynamic disciplines in the development of information technology.

[0004] CN116091604A discloses a method for locating drainage wire clamps for live working in distribution networks based on computer vision, which includes the following steps: grouping the collected images, constructing a DeepLabv3+ semantic segmentation convolutional neural network model, processing the collected images through image enhancement technology, dividing them into a training set, a test set and a validation set, training the DeepLabv3+ semantic segmentation convolutional neural network model using a five-fold cross-validation method, and using the trained model to locate and identify the central screw of the collected drainage wire clamp image.

[0005] At present, a visual monitor is generally used to realize the visual monitoring operation of the set shelf placement scene in the shopping mall to obtain the corresponding shelf placement image, and then the shelf placement image is filtered to obtain the corresponding monitoring screen. However, the data content of each shelf placement image corresponding to different times is different. If the same filtering algorithm is used, it will inevitably lead to that the filtered pictures of each shelf placement image are not all optimal. Therefore, a dynamic filtering processing mode is needed to provide the best filtering algorithm required by each shelf placement image.

[0006] The invention application publication CN113947351A discloses a material monitoring system. By setting an information reading module, a route detection module, a shelf monitoring module and an image monitoring module, for a warehouse with a shelf area and a stacking area, the material information is first read when the material is put into the warehouse, and then according to the detection result of the route detection module, it is judged whether the material is transported to the shelf area or the stacking area. After the judgment is completed, the shelf monitoring module and the image monitoring module are used to detect the material changes in the corresponding shelf area and the stacking area, and the information of the materials entering and leaving the shelf area is obtained, so that the material entry and exit of the warehouse can be monitored. Invention application publication CN110557601A discloses a shelf monitoring system and method, including at least one straight-line distance sensor: used to detect whether there is an obstruction at the pickup port of the shelf; an image acquisition device: used to collect image information in front of the shelf; a control device: used to control the image acquisition device to start collecting image information in front of the shelf when the straight-line distance sensor detects that there is an obstruction at the pickup port of the shelf, until the straight-line distance sensor detects that the obstruction is separated from the pickup port of the shelf, and control the image acquisition device to stop collecting image information in front of the shelf; when it is detected that there is an obstruction at the pickup port of the shelf, start collecting image information in front of the shelf, until it is detected that the obstruction is separated from the pickup port of the shelf, and stop collecting image information in front of the shelf, thereby reducing the collection and storage of image information not related to picking up goods on the shelf and reducing the power consumption of the system. Summary of the invention

[0007] In order to overcome the technical problems in the prior art, the present invention proposes a computer image processing system based on an artificial intelligence model, which can input multiple information parameters of each distribution area, the clarity of the cubic interpolation image, the number of edge pixels and the algorithm identifier corresponding to the current filtering algorithm into the AI ​​prediction model and run the AI ​​prediction model to obtain the maximum noise amplitude of the picture content obtained after the current filtering algorithm is executed on the cubic interpolation image output by the AI ​​prediction model, and numerically compare the maximum noise amplitudes corresponding to various filtering algorithms respectively, and use the filtering algorithm corresponding to the maximum noise amplitude with the smallest numerical value to perform filtering operation on the cubic interpolation image obtained after targeted enhancement processing of the on-site image to obtain an optimized processed image, and the optimized processed image is used to be displayed at the remote monitoring server corresponding to the shopping mall to improve the monitoring effect of the shelf placement scenes in various places in the shopping mall.

[0008] The system comprises: A time analysis device, used to analyze whether the current time is within the operation time period of the mall, and to send a first analysis signal when the current time is within the operation time period of the mall, and to send a second analysis signal when the current time is not within the operation time period of the mall; A visual monitor connected to the time analysis device, for implementing a visual monitoring operation of a set shelf placement scene in a shopping mall when the first analysis signal is received, so as to obtain a corresponding shelf placement image; The order mapping device comprises a first processing device, a second processing device and a third processing device, wherein the second processing device is connected to the first processing device and the third processing device respectively, the first processing device is connected to the visual monitor and is used to perform a non-scaling transformation blur process on the received shelf placement image to obtain and output a corresponding smoothed image, the second processing device is used to perform a Butterworth low-pass filtering process on the received smoothed image to obtain and output a corresponding low-pass filtered image, and the third processing device is used to perform a cubic polynomial interpolation process on the received low-pass filtered image to obtain and output a corresponding cubic interpolation image; A content identification mechanism, connected to the order mapping device, for identifying each edge pixel point in the received cubic interpolation image, and combining the each edge pixel point to obtain each distribution area in the cubic interpolation image; A region processing mechanism connected to the content identification mechanism, used to detect multiple information parameters of each distribution region to obtain multiple information parameters of each distribution region, wherein the multiple information parameters of each distribution region are color component values ​​corresponding to each pixel point in the partitioned region, and the color component value corresponding to each pixel point includes a red and green component value, a black and white component value, and a yellow and blue component value corresponding to the pixel point; A directional prediction mechanism, connected to the regional processing mechanism, for inputting a plurality of information parameters of each distribution area, the clarity of the cubic interpolation image, the number of edge pixels, and an algorithm identifier corresponding to the current filtering algorithm into an AI prediction model and running the AI ​​prediction model to obtain the maximum noise amplitude of the picture content obtained after the current filtering algorithm is executed on the cubic interpolation image output by the AI ​​prediction model, wherein the AI ​​prediction model is a feedforward deep network after completing each training; An optimization processing mechanism, connected to the directional prediction mechanism, is used to compare the maximum noise amplitudes corresponding to various filtering algorithms, and use the filtering algorithm corresponding to the maximum noise amplitude with the smallest value to perform a filtering operation on the cubic interpolation image to obtain an optimized processing image; The data playback mechanism is arranged at the remote monitoring server corresponding to the shopping mall and connected to the optimization processing mechanism, and is used for receiving and displaying the optimization processing image.

[0009] The computer image processing system based on the artificial intelligence model of the present invention has compact logic and intelligent operation. Since the maximum noise amplitude of the picture content obtained after the current filtering algorithm can be intelligently analyzed based on the image data and the algorithm identifier corresponding to the current filtering algorithm, the filtering algorithm corresponding to the minimum value of each maximum noise amplitude corresponding to each filtering algorithm is used as the target filtering algorithm, thereby improving the monitoring effect of the shopping mall shelf placement scene. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The embodiments of the present invention will be described below with reference to the accompanying drawings.

[0011] Figure 1 It is a block diagram of the internal structure of a computer image processing system based on an artificial intelligence model according to Embodiment 1 of the present invention.

[0012] Figure 2 It is a block diagram of the internal structure of a computer image processing system based on an artificial intelligence model according to Embodiment 2 of the present invention.

[0013] Figure 3 It is a block diagram of the internal structure of a computer image processing system based on an artificial intelligence model according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0014] The implementation scheme of the computer image processing system based on the artificial intelligence model of the present invention will be described in detail below with reference to the accompanying drawings.

[0015] Implementation 1 Figure 1 The internal structure block diagram of the computer image processing system based on the artificial intelligence model according to Embodiment 1 of the present invention is shown, and the system includes: A time analysis device, used to analyze whether the current time is within the operation time period of the mall, and to send a first analysis signal when the current time is within the operation time period of the mall, and to send a second analysis signal when the current time is not within the operation time period of the mall; For example, the time analysis device is used to analyze whether the current time is within the operation time period of the mall, and when the current time is within the operation time period of the mall, sends a first analysis signal, and when the current time is not within the operation time period of the mall, sends a second analysis signal, including: a CPLD device can be selected to implement the time analysis device, which is used to analyze whether the current time is within the operation time period of the mall, and when the current time is within the operation time period of the mall, sends a first analysis signal, and when the current time is not within the operation time period of the mall, sends a second analysis signal; A visual monitor connected to the time analysis device, for implementing a visual monitoring operation of a set shelf placement scene in a shopping mall when the first analysis signal is received, so as to obtain a corresponding shelf placement image; The order mapping device comprises a first processing device, a second processing device and a third processing device, wherein the second processing device is connected to the first processing device and the third processing device respectively, the first processing device is connected to the visual monitor and is used to perform a non-scaling transformation blur process on the received shelf placement image to obtain and output a corresponding smoothed image, the second processing device is used to perform a Butterworth low-pass filtering process on the received smoothed image to obtain and output a corresponding low-pass filtered image, and the third processing device is used to perform a cubic polynomial interpolation process on the received low-pass filtered image to obtain and output a corresponding cubic interpolation image; A content identification mechanism, connected to the order mapping device, for identifying each edge pixel point in the received cubic interpolation image, and combining the each edge pixel point to obtain each distribution area in the cubic interpolation image; A region processing mechanism connected to the content identification mechanism, used to detect multiple information parameters of each distribution region to obtain multiple information parameters of each distribution region, wherein the multiple information parameters of each distribution region are color component values ​​corresponding to each pixel point in the partitioned region, and the color component value corresponding to each pixel point includes a red and green component value, a black and white component value, and a yellow and blue component value corresponding to the pixel point; A directional prediction mechanism, connected to the regional processing mechanism, for inputting a plurality of information parameters of each distribution area, the clarity of the cubic interpolation image, the number of edge pixels, and an algorithm identifier corresponding to the current filtering algorithm into an AI prediction model and running the AI ​​prediction model to obtain the maximum noise amplitude of the picture content obtained after the current filtering algorithm is executed on the cubic interpolation image output by the AI ​​prediction model, wherein the AI ​​prediction model is a feedforward deep network after completing each training; An optimization processing mechanism, connected to the directional prediction mechanism, is used to compare the maximum noise amplitudes corresponding to various filtering algorithms, and use the filtering algorithm corresponding to the maximum noise amplitude with the smallest value to perform a filtering operation on the cubic interpolation image to obtain an optimized processing image; A data playback mechanism, which is arranged at a remote monitoring server corresponding to the shopping mall and connected to the optimization processing mechanism, and is used to receive and display the optimization processing image; Among them, various filtering algorithms include statistical sorting filtering algorithm, maximum filtering algorithm, minimum filtering algorithm, midpoint filtering algorithm, alpha mean filtering algorithm, adaptive filtering algorithm, bilateral filtering algorithm, guided filtering algorithm and box filtering algorithm; The visual monitor is further configured to terminate the visual monitoring operation on the placement scene of the set shelves in the shopping mall when the second analysis signal is received.

[0016] Implementation 2 Figure 2 It is a block diagram of the internal structure of a computer image processing system based on an artificial intelligence model according to Embodiment 2 of the present invention.

[0017] Compared to Figure 1 , Figure 2 The computer image processing system based on the artificial intelligence model may also include the following components: A humidity measuring mechanism, comprising a plurality of humidity measuring units, for respectively measuring current surface humidity values ​​of the content identification mechanism, the region processing mechanism, the first processing device, the second processing device, and the third processing device; The humidity measuring mechanism comprises a plurality of humidity measuring units for respectively measuring the current surface humidity values ​​of the content identification mechanism, the region processing mechanism, the first processing device, the second processing device and the third processing device, comprising: the plurality of humidity measuring units respectively used for the content identification mechanism, the region processing mechanism, the first processing device, the second processing device and the third processing device are a plurality of non-contact humidity sensors; The humidity measuring mechanism includes a plurality of humidity measuring units for respectively measuring the current surface humidity values ​​of the content identification mechanism, the region processing mechanism, the first processing device, the second processing device and the third processing device, and further includes: the internal structures of the plurality of non-contact humidity sensors respectively used for the content identification mechanism, the region processing mechanism, the first processing device, the second processing device and the third processing device are the same; The humidity measuring mechanism includes a plurality of humidity measuring units for respectively measuring the current surface humidity values ​​of the content identification mechanism, the region processing mechanism, the first processing device, the second processing device and the third processing device, and further includes: a plurality of non-contact humidity sensors respectively used for the content identification mechanism, the region processing mechanism, the first processing device, the second processing device and the third processing device have the same humidity measurement upper limit threshold and humidity measurement lower limit threshold; Among them, the humidity measuring mechanism includes multiple humidity measuring units, which are used to respectively measure the current surface humidity values ​​of the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device. It also includes: multiple non-contact humidity sensors respectively used for the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device are at equal distances from the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device.

[0018] Implementation 3 Figure 3 It is a block diagram of the internal structure of a computer image processing system based on an artificial intelligence model according to Embodiment 3 of the present invention.

[0019] Compared to Figure 1 , Figure 3 The computer image processing system based on the artificial intelligence model may also include the following components: an immediate notification mechanism, connected to the plurality of non-contact humidity sensors respectively used by the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device, and the third processing device, for executing corresponding humidity alarm actions based on humidity measurement results of the plurality of non-contact humidity sensors respectively used by the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device, and the third processing device; The instant notification mechanism is respectively connected to a plurality of non-contact humidity sensors respectively used by the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, and is used to perform corresponding humidity alarm actions based on humidity measurement results of the plurality of non-contact humidity sensors respectively used by the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, including: a humidity receiving unit, a humidity judgment unit and a notification execution unit built into the instant notification mechanism; Among them, the instant notification mechanism is respectively connected to multiple non-contact humidity sensors used by the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, and is used to perform corresponding humidity alarm actions based on the humidity measurement results of the multiple non-contact humidity sensors used by the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, and also includes: in the instant notification mechanism, the humidity receiving unit, the humidity judgment unit and the notification execution unit are connected in sequence.

[0020] In addition, in the computer image processing system based on the artificial intelligence model, the second processing device is connected to the first processing device and the third processing device respectively, the first processing device is connected to the visual monitor, and is used to perform non-scaling transformation blur processing on the received shelf placement image to obtain and output a corresponding smoothed image, the second processing device is used to perform Butterworth low-pass filtering processing on the received smoothed image to obtain and output a corresponding low-pass filtered image, and the third processing device is used to perform cubic polynomial interpolation processing on the received low-pass filtered image to obtain and output a corresponding cubic interpolation image, including: using different ASIC chips to implement the second processing device, the first processing device and the third processing device respectively.

[0021] Beneficial technical effects of the present invention: 1. Identify each edge pixel point in the received cubic interpolation image, combine the each edge pixel point to obtain each distribution area in the cubic interpolation image, detect multiple information parameters of each distribution area to obtain multiple information parameters of each distribution area, the multiple information parameters of each distribution area are color component values ​​corresponding to each pixel point in the partition area, and the color component value corresponding to each pixel point includes a red and green component value, a black and white component value, and a yellow and blue component value corresponding to the pixel point; Second: input multiple information parameters of each distribution area, the clarity of the cubic interpolation image, the number of edge pixels, and the algorithm identifier corresponding to the current filtering algorithm into the AI ​​prediction model and run the AI ​​prediction model to obtain the maximum noise amplitude of the picture content obtained after the current filtering algorithm is executed on the cubic interpolation image output by the AI ​​prediction model, wherein the AI ​​prediction model is a feedforward deep network after completing each training; 3: Numerically compare the maximum noise amplitudes corresponding to various filtering algorithms, and use the filtering algorithm corresponding to the maximum noise amplitude with the smallest numerical value to perform filtering operation on the cubic interpolation image obtained after targeted enhancement processing of the on-site image to obtain an optimized processed image. The optimized processed image is used to be displayed at the remote monitoring server corresponding to the shopping mall to improve the monitoring effect of the shelf placement scenes in various places in the shopping mall.

[0022] Although the above description may be considered as the best mode and / or other examples, it should be understood that various modifications may be made herein and the subject matter disclosed herein may be implemented in various forms and examples, and the teachings may be applied in numerous applications, of which only some are disclosed herein. The following claims are intended to claim any application, modification, and variation that falls within the true scope of the present teachings.

Claims

1. A computer image processing system based on an artificial intelligence model, characterized in that: The system comprises: A time analysis device, used to analyze whether the current time is within the operation time period of the mall, and to send a first analysis signal when the current time is within the operation time period of the mall, and to send a second analysis signal when the current time is not within the operation time period of the mall; A visual monitor connected to the time analysis device, for implementing a visual monitoring operation of a set shelf placement scene in a shopping mall when the first analysis signal is received, so as to obtain a corresponding shelf placement image; The order mapping device comprises a first processing device, a second processing device and a third processing device, wherein the second processing device is connected to the first processing device and the third processing device respectively, the first processing device is connected to the visual monitor and is used to perform a non-scaling transformation blur process on the received shelf placement image to obtain and output a corresponding smoothed image, the second processing device is used to perform a Butterworth low-pass filtering process on the received smoothed image to obtain and output a corresponding low-pass filtered image, and the third processing device is used to perform a cubic polynomial interpolation process on the received low-pass filtered image to obtain and output a corresponding cubic interpolation image; A content identification mechanism, connected to the order mapping device, for identifying each edge pixel point in the received cubic interpolation image, and combining the each edge pixel point to obtain each distribution area in the cubic interpolation image; A region processing mechanism connected to the content identification mechanism, used to detect multiple information parameters of each distribution region to obtain multiple information parameters of each distribution region, wherein the multiple information parameters of each distribution region are color component values ​​corresponding to each pixel point in the partitioned region, and the color component value corresponding to each pixel point includes a red and green component value, a black and white component value, and a yellow and blue component value corresponding to the pixel point; A directional prediction mechanism, connected to the regional processing mechanism, for inputting a plurality of information parameters of each distribution area, the clarity of the cubic interpolation image, the number of edge pixels, and an algorithm identifier corresponding to the current filtering algorithm into an AI prediction model and running the AI ​​prediction model to obtain the maximum noise amplitude of the picture content obtained after the current filtering algorithm is executed on the cubic interpolation image output by the AI ​​prediction model, wherein the AI ​​prediction model is a feedforward deep network after completing each training; An optimization processing mechanism, connected to the directional prediction mechanism, is used to compare the maximum noise amplitudes corresponding to various filtering algorithms, and use the filtering algorithm corresponding to the maximum noise amplitude with the smallest value to perform a filtering operation on the cubic interpolation image to obtain an optimized processing image; The data playback mechanism is arranged at the remote monitoring server corresponding to the shopping mall and connected to the optimization processing mechanism, and is used for receiving and displaying the optimization processing image.

2. The computer image processing system based on the artificial intelligence model as claimed in claim 1, characterized in that: Various filtering algorithms include statistical sorting filtering algorithm, maximum filtering algorithm, minimum filtering algorithm, midpoint filtering algorithm, alpha mean filtering algorithm, adaptive filtering algorithm, bilateral filtering algorithm, guided filtering algorithm and box filtering algorithm; The visual monitor is further configured to terminate the visual monitoring operation on the placement scene of the set shelves in the shopping mall when the second analysis signal is received.

3. The computer image processing system based on the artificial intelligence model as claimed in claim 2, characterized in that: The system further comprises: A humidity measuring mechanism, comprising a plurality of humidity measuring units, for respectively measuring current surface humidity values ​​of the content identification mechanism, the region processing mechanism, the first processing device, the second processing device, and the third processing device; Among them, the humidity measuring mechanism includes multiple humidity measuring units, which are used to respectively measure the current surface humidity values ​​of the content identification mechanism, the area processing mechanism, the first processing device, the second processing device and the third processing device, including: the multiple humidity measuring units respectively used for the content identification mechanism, the area processing mechanism, the first processing device, the second processing device and the third processing device are multiple non-contact humidity sensors.

4. The computer image processing system based on the artificial intelligence model as claimed in claim 3, characterized in that: The humidity measuring mechanism includes a plurality of humidity measuring units for respectively measuring the current surface humidity values ​​of the content identification mechanism, the area processing mechanism, the first processing device, the second processing device and the third processing device. The mechanism also includes: the internal structures of the plurality of non-contact humidity sensors respectively used for the content identification mechanism, the area processing mechanism, the first processing device, the second processing device and the third processing device are the same.

5. The computer image processing system based on an artificial intelligence model as claimed in claim 3, characterized in that: The humidity measuring mechanism includes a plurality of humidity measuring units for respectively measuring the current surface humidity values ​​of the content identification mechanism, the area processing mechanism, the first processing device, the second processing device and the third processing device, and further includes: a plurality of non-contact humidity sensors respectively used for the content identification mechanism, the area processing mechanism, the first processing device, the second processing device and the third processing device have the same humidity measurement upper limit threshold and humidity measurement lower limit threshold.

6. The computer image processing system based on the artificial intelligence model as claimed in claim 5, characterized in that: The humidity measuring mechanism includes a plurality of humidity measuring units for respectively measuring the current surface humidity values ​​of the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, and further includes: a plurality of non-contact humidity sensors respectively used for the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device are at equal distances from the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device.

7. The computer image processing system based on an artificial intelligence model according to any one of claims 3 to 6, characterized in that: The system further comprises: The instant notification mechanism is respectively connected to the multiple non-contact humidity sensors respectively used for the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, and is used to perform corresponding humidity alarm actions based on the humidity measurement results of the multiple non-contact humidity sensors respectively used for the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device.

8. The computer image processing system based on the artificial intelligence model as claimed in claim 7, characterized in that: The instant notification mechanism is respectively connected to a plurality of non-contact humidity sensors respectively used for the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, and is used to perform corresponding humidity alarm actions based on the humidity measurement results of the plurality of non-contact humidity sensors respectively used for the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, including: the instant notification mechanism has a built-in humidity receiving unit, a humidity judgment unit and a notification execution unit.

9. The computer image processing system based on the artificial intelligence model as claimed in claim 8, characterized in that: An instant notification mechanism is respectively connected to a plurality of non-contact humidity sensors respectively used for the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device, and is used to perform corresponding humidity alarm actions based on the humidity measurement results of the plurality of non-contact humidity sensors respectively used for the content identification mechanism, the regional processing mechanism, the first processing device, the second processing device and the third processing device. It also includes: within the instant notification mechanism, the humidity receiving unit, the humidity judgment unit and the notification execution unit are connected in sequence.

Images