Fixed transportation environment visual content management system

Abstract

This invention relates to a visual content management system for fixed traffic environments, comprising: a directional acquisition device fixed above a train station platform to acquire and output a currently captured image; and a sequence resolution device installed at the visual monitoring end, used to intelligently identify, using an Extreme Learning Machine neural network model, the binary value representing the order of image processing performed on the currently captured image within the directional acquisition device, based on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image. Through this invention, the binary value representing the order of image processing performed on the currently captured image within the directional acquisition device can be intelligently identified using an Extreme Learning Machine neural network model based on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image, thereby clarifying the order of various image processing performed on the currently captured image and assisting the monitoring end in performing the restoration processing of the original image.

Description

Technical Field

[0001] This invention relates to the field of traffic management, and more particularly to a visual content management system for fixed traffic environments. Background Technology

[0002] The purpose of traffic management is to understand and follow the inherent objective laws of road traffic flow, and to continuously improve the efficiency and quality of traffic management by using modern technology and scientific principles, methods and measures, so as to achieve less delay, shorter travel time, greater traffic capacity, better order and lower operating costs, thereby obtaining the best social, economic, transportation and environmental benefits, and serving the economic development and the improvement of people's living standards and travel quality.

[0003] In traffic management, trains are a primary mode of transportation. Generally, monitoring devices can be fixed above the train station platform to perform visual monitoring operations on the platform when the train is stopped, in order to acquire and output corresponding captured images. The problem is that even if the image receiving end knows that the received image has undergone various image processing steps, it is still difficult to reconstruct the original image that reflects the actual scene of the train station platform because the processing order of each step cannot be determined. Summary of the Invention

[0004] To address the technical problems in existing technologies, this invention provides a fixed traffic environment visual content management system. Based on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image, an extreme learning machine neural network model intelligently identifies binary values ​​representing the order of image processing performed on the currently captured image within the directional acquisition device. These binary values ​​include sequential values ​​corresponding to each image processing step, with each step represented by its execution sequence number plus its ASCII code. This clarifies the order of the various image processing steps performed on the currently captured image. The various shooting parameters of the directional acquisition device include the shutter speed, exposure, and resolution of the directional acquisition device. The various visualization data of the currently captured image include the maximum noise amplitude, the number of noise types, the background area ratio, and the average brightness values ​​of each pixel. This provides key basic information for the sequential analysis of various image processing. The structure of the Extreme Learning Machine (ELM) neural network model used to perform the sequential analysis of various image processing is a specially designed structure. Specifically, a training mapping device is introduced to perform multiple training actions on the ELM neural network to obtain the ELM neural network after multiple training actions, which is then used as the output of the ELM neural network model.

[0005] According to the present invention, a visual content management system for fixed traffic environments is provided, the system comprising:

[0006] A directional acquisition device is fixed above the train station platform to perform visual monitoring operations facing the train station platform when the train is stopped, in order to acquire and output the corresponding currently captured images;

[0007] A wireless notification device, connected to a sequence resolution device, is used to notify a remote blockchain processing node via a wireless data channel the received binary value representing the order of various image processing operations performed on the currently captured image in the directional acquisition device.

[0008] The training mapping device is used to perform multiple training actions on the Extreme Learning Machine Neural Network to obtain the Extreme Learning Machine Neural Network after completing multiple training actions and output it as the Extreme Learning Machine Neural Network model.

[0009] The sequence resolution device is installed at the visual monitoring end and connected to the directional acquisition device, the wireless notification device, and the training mapping device respectively. It is used to intelligently identify the binary value representing the order of various image processing performed on the directional acquisition device in the current captured image using an extreme learning machine neural network model based on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image. The binary value includes the sequence values ​​corresponding to each image processing connected end to end. The sequence value corresponding to each image processing is represented by the execution sequence number of the image processing plus the ASCII code of the image processing.

[0010] Among them, the binary values ​​representing the order of image processing performed on the current captured image in the directional acquisition device are intelligently identified using an extreme learning machine neural network model based on the various shooting parameters of the directional acquisition device and the various visualization data of the current captured image. These include: the various shooting parameters of the directional acquisition device are the shutter speed, exposure, and resolution of the directional acquisition device;

[0011] Among them, the binary values ​​representing the order of image processing performed on the current captured image in the directional acquisition device, which are based on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image, are intelligently identified using the Extreme Learning Machine neural network model. These values ​​also include the following visualization data of the current captured image: the maximum noise amplitude, the number of noise types, the background area ratio, and the average of the brightness values ​​corresponding to each pixel.

[0012] The binary value includes the sequential values ​​corresponding to each image processing step, which are connected end to end. The sequential value corresponding to each image processing step is represented by the execution sequence number of the image processing step plus the ASCII code of the image processing step. Each image processing step is one of the following: point image restoration processing, rotation correction processing, image sharpening processing, image enhancement processing, and image filtering processing.

[0013] Therefore, it can be seen that the present invention has at least the following main inventive concepts:

[0014] The first inventive concept: Based on the various shooting parameters of the directional acquisition device and the various visualization data of the currently captured image, an extreme learning machine neural network model is used to intelligently identify the binary value representing the order of various image processing performed on the currently captured image in the directional acquisition device. The binary value includes the order value corresponding to each image processing connected end to end. The order value corresponding to each image processing is represented by the execution sequence number of the image processing plus the ASCII code of the image processing, thereby clarifying the order of various image processing performed on the currently captured image.

[0015] The second inventive concept: the various shooting parameters of the directional acquisition device are the shutter speed, exposure and resolution of the directional acquisition device, and the various visual data of the currently captured image are the maximum noise amplitude, the number of noise types, the background area ratio and the average value of each brightness value corresponding to each pixel, thereby providing key basic information for the sequential analysis of various image processing.

[0016] The third inventive concept: The structure of the Extreme Learning Machine Neural Network Model used for performing sequence analysis of various image processing is a targeted design structure. Specifically, a training mapping device is introduced to perform multiple training actions on the Extreme Learning Machine Neural Network to obtain the Extreme Learning Machine Neural Network after completing multiple training actions, which is then used as the output of the Extreme Learning Machine Neural Network Model. Attached Figure Description

[0017] The embodiments of the present invention will now be described with reference to the accompanying drawings, wherein:

[0018] Figure 1 This is a structural block diagram of a visual content management system for a fixed traffic environment according to a primary embodiment of the present invention.

[0019] Figure 2 This is a structural block diagram of a visual content management system for a fixed traffic environment according to a secondary embodiment of the present invention.

[0020] Figure 3 This is a structural block diagram of a visual content management system for a fixed traffic environment according to a further embodiment of the present invention. Detailed Implementation

[0021] The embodiments of the fixed traffic environment visual content management system of the present invention will now be described in detail with reference to the accompanying drawings.

[0022] Figure 1 The structural block diagram of a visual content management system for fixed traffic environments, as shown in the primary embodiment of the present invention, includes:

[0023] A directional acquisition device is fixed above the train station platform to perform visual monitoring operations facing the train station platform when the train is stopped, in order to acquire and output the corresponding currently captured images;

[0024] A wireless notification device, connected to a sequence resolution device, is used to notify a remote blockchain processing node via a wireless data channel the received binary value representing the order of various image processing operations performed on the currently captured image in the directional acquisition device.

[0025] Specifically, the wireless notification device, connected to the order parsing device, is used to notify the remote blockchain processing node of the received binary value representing the order of various image processing operations performed on the currently captured image in the directional acquisition device via a wireless data channel, including: the wireless data channel being a time-division duplex communication link or a frequency-division duplex communication link.

[0026] The training mapping device is used to perform multiple training actions on the Extreme Learning Machine Neural Network to obtain the Extreme Learning Machine Neural Network after completing multiple training actions and output it as the Extreme Learning Machine Neural Network model.

[0027] The sequence resolution device is installed at the visual monitoring end and connected to the directional acquisition device, the wireless notification device, and the training mapping device respectively. It is used to intelligently identify the binary value representing the order of various image processing performed on the directional acquisition device in the current captured image using an extreme learning machine neural network model based on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image. The binary value includes the sequence values ​​corresponding to each image processing connected end to end. The sequence value corresponding to each image processing is represented by the execution sequence number of the image processing plus the ASCII code of the image processing.

[0028] Among them, the binary values ​​representing the order of image processing performed on the current captured image in the directional acquisition device are intelligently identified using an extreme learning machine neural network model based on the various shooting parameters of the directional acquisition device and the various visualization data of the current captured image. These include: the various shooting parameters of the directional acquisition device are the shutter speed, exposure, and resolution of the directional acquisition device;

[0029] Among them, the binary values ​​representing the order of image processing performed on the current captured image in the directional acquisition device, which are based on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image, are intelligently identified using the Extreme Learning Machine neural network model. These values ​​also include the following visualization data of the current captured image: the maximum noise amplitude, the number of noise types, the background area ratio, and the average of the brightness values ​​corresponding to each pixel.

[0030] The binary value includes the sequential values ​​of each image processing step connected end to end. The sequential value of each image processing step is represented by the execution sequence number of the image processing step plus the ASCII code of the image processing step. Each image processing step is one of the following: point image restoration processing, rotation correction processing, image sharpening processing, image enhancement processing, and image filtering processing.

[0031] Furthermore, based on the various shooting parameters of the directional acquisition device and the various visualization data of the currently captured image, the Extreme Learning Machine neural network model intelligently identifies the binary value representing the order of various image processing performed on the directional acquisition device for the currently captured image. The binary value includes the order values ​​corresponding to each image processing connected end to end. The order value corresponding to each image processing is represented by the execution sequence number of the image processing plus the ASCII code of the image processing. This includes: inputting the various shooting parameters of the directional acquisition device and the various visualization data of the currently captured image into the Extreme Learning Machine neural network model in parallel.

[0032] Figure 2 This is a structural block diagram of a visual content management system for a fixed traffic environment according to a secondary embodiment of the present invention.

[0033] Compared to Figure 1 , Figure 2 The fixed traffic environment visual content management system may also include:

[0034] An ASIC processing device is connected to the order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device, respectively, and is used to provide configuration operations of working parameters for the order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device in a time-division manner.

[0035] The ASIC processing device is connected to the order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device respectively, and is used to provide configuration operations for working parameters to the order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device in a time-division manner. This includes the order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device sharing the same working parameter configuration interface.

[0036] The order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device use different configuration address data.

[0037] Parallel connections are established between each pair of the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device via a parallel data bus;

[0038] The parallel connection established between each pair of the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device via a parallel data bus includes: the parallel data bus being one of an 8-bit parallel data bus, a 16-bit parallel data bus, and a 32-bit parallel data bus.

[0039] Figure 3 This is a structural block diagram of a visual content management system for a fixed traffic environment according to a further embodiment of the present invention.

[0040] Compared to Figure 1 , Figure 3 The fixed traffic environment visual content management system may also include:

[0041] Power support devices are connected to the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device respectively, and are used to provide power distribution support with different operating voltages to the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device in a time-division manner.

[0042] The power support device is connected to the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device respectively, and is used to provide power distribution support with different operating voltages for the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device in a time-division manner. This includes: two or more devices among the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device that have the same operating voltage requirement using the same power supply line.

[0043] The power support device, which is connected to the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device respectively, and is used to provide power distribution support with different operating voltages for the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device in a time-division manner, further includes: the power support device is an uninterruptible power supply device.

[0044] The power support device, which is connected to the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device respectively, and is used to provide power distribution support for the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device at different times, further includes providing different operating voltages, including a 3.3V operating voltage, for the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device at different times.

[0045] The power support device, which is connected to the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device respectively, and is used to provide power distribution support for the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device at different times, further includes providing different operating voltages, including a 5V operating voltage, for the sequence resolution device, the orientation acquisition device, the wireless notification device, and the training mapping device at different times.

[0046] In addition, in the fixed traffic environment visual content management system, the parallel input of various shooting parameters of the directional acquisition device and various visualization data of the currently captured image into the extreme learning machine neural network model includes: performing binary numerical conversion processing on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image respectively before inputting them into the extreme learning machine neural network model in parallel; and using a parallel control interface to execute the data processing process of performing binary numerical conversion processing on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image respectively before inputting them into the extreme learning machine neural network model in parallel.

[0047] The process of performing binary numerical conversion on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image before inputting them in parallel into the Extreme Learning Machine neural network model, and using a parallel control interface to perform binary numerical conversion on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image before inputting them in parallel into the Extreme Learning Machine neural network model, includes: selecting to use a CPLD chip to implement the parallel control interface.

[0048] The fixed traffic environment visual content management system of this invention addresses the technical problem in existing technologies that make it difficult to identify the true content of received traffic scene images. By using an extreme learning machine neural network model based on various shooting parameters of the directional acquisition device and various visualization data of the currently captured image, it intelligently identifies the binary values ​​representing the order of various image processing operations performed on the currently captured image in the directional acquisition device. This clarifies the order of various image processing operations performed on the currently captured image, helping the monitoring end to perform the restoration processing of the original image, thereby solving the aforementioned technical problem.

[0049] Although the invention has been described with respect to a limited number of embodiments, those skilled in the art will understand from this disclosure that other embodiments can be devised without departing from the scope of the invention disclosed herein. Therefore, the scope of the invention should only be limited by the appended claims.

Claims

1. A visual content management system for fixed traffic environments, characterized in that, The system includes: A directional acquisition device is fixed above the train station platform to perform visual monitoring operations facing the train station platform when the train is stopped, in order to acquire and output the corresponding currently captured images; A wireless notification device, connected to a sequence resolution device, is used to notify a remote blockchain processing node via a wireless data channel the received binary value representing the order of various image processing operations performed on the currently captured image in the directional acquisition device. The training mapping device is used to perform multiple training actions on the Extreme Learning Machine Neural Network to obtain the Extreme Learning Machine Neural Network after completing multiple training actions and output it as the Extreme Learning Machine Neural Network model. The sequence resolution device is installed at the visual monitoring end and connected to the directional acquisition device, the wireless notification device, and the training mapping device. It is used to intelligently identify the binary value representing the order of various image processing operations performed on the directional acquisition device in the current captured image based on the various shooting parameters of the directional acquisition device and the various visualization data of the currently captured image using an extreme learning machine neural network model. The binary value includes the sequence values ​​of each image processing operation connected end to end. The sequence value of each image processing operation is represented by the execution sequence number of the image processing operation plus the ASCII code of the image processing operation. Among them, the shooting parameters of the directional acquisition device are the shutter speed, exposure, and resolution of the directional acquisition device; Among them, the various visualization data of the currently captured image are the maximum noise amplitude, the number of noise types, the background area ratio, and the average of the brightness values ​​corresponding to each pixel. Each image processing step includes one of the following: point image restoration processing, rotation correction processing, image sharpening processing, image enhancement processing, and image filtering processing. In this process, various shooting parameters of the directional acquisition device and various visualization data of the currently captured image are input in parallel into the extreme learning machine neural network model; The process involves performing binary numerical conversion on the various shooting parameters of the directional acquisition device and the various visualization data of the currently captured image, and then inputting them in parallel into the Extreme Learning Machine neural network model. It also involves using a parallel control interface to perform the data processing process of performing binary numerical conversion on the various shooting parameters of the directional acquisition device and the various visualization data of the currently captured image, and then inputting them in parallel into the Extreme Learning Machine neural network model. We chose to use a CPLD chip to implement the parallel control interface.

2. The fixed traffic environment visual content management system as described in claim 1, characterized in that, The system also includes: An ASIC processing device is connected to the order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device, respectively, and is used to provide configuration operations of working parameters for the order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device in a time-division manner. The order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device share the same working parameter configuration interface.

3. The fixed traffic environment visual content management system as described in claim 2, characterized in that: The order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device use different configuration address data.

4. The fixed traffic environment visual content management system as described in claim 2, characterized in that: Parallel connections for communication data links are established between each pair of the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device via a parallel data bus. The parallel data bus is one of an 8-bit parallel data bus, a 16-bit parallel data bus, and a 32-bit parallel data bus.

5. The fixed traffic environment visual content management system as described in claim 1, characterized in that, The system also includes: Power support devices are connected to the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device respectively, and are used to provide power distribution support with different operating voltages to the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device in a time-division manner. Among them, in the order parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device, two or more devices with the same operating voltage requirement use the same power supply line.

6. The fixed traffic environment visual content management system as described in claim 5, characterized in that: The power support device is an uninterruptible power supply (UPS).

7. The fixed traffic environment visual content management system as described in claim 6, characterized in that: The sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device are provided with different operating voltages, including a 3.3V operating voltage, in a time-division manner.

8. The fixed traffic environment visual content management system as described in claim 7, characterized in that: The time-division multiplexing mechanism provides different operating voltages, including a 5V operating voltage, to the sequence parsing device, the orientation acquisition device, the wireless notification device, and the training mapping device.

Images