Real-time Generated Picture Information Extraction System

Abstract

The present invention relates to a real-time generated picture information extraction system, comprising: a timing acquisition mechanism for acquiring real-time generated pictures corresponding to the internal environment of a school; an information conversion device for parallelly inputting each content data of the standard original images of the timing acquisition mechanism and multiple visualization data of the real-time generated pictures into an artificial intelligence model to obtain a combination of image processing types through which the RAW image corresponding to the real-time generated picture output by the artificial intelligence model reaches the real-time generated picture. Through the present invention, it is possible to intelligently identify, based on an artificially intelligent model designed specifically for the purpose and various visualization data obtained through customized screening, the combination of image processing types through which the RAW image corresponding to the real-time generated picture reaches the real-time generated picture, and the combination of image processing types is represented by sequentially connecting multiple type names respectively corresponding to multiple image processing types, thereby helping the monitoring end to clarify the true content of the real-time generated picture.

Description

Technical Field

[0001] The present invention relates to the field of signal analysis, and particularly to a real-time generated picture information extraction system. Background Art

[0002] Signal analysis can be understood as decomposing a composite signal into the accumulation of n simple signal components, that is, grasping the main components of the signal to simplify the complex, so as to further analyze the characteristics of these signal components to determine the characteristics of the composite signal. The commonly used signal analysis method is to use frequency as the independent variable and perform spectral analysis in the frequency domain dimension. For example, a three-dimensional spectrum diagram can be used to reflect the relationship between the amplitudes and frequencies (or time) of the components of a signal; among them, the change of the time-domain waveform can be observed from the time-amplitude dimension, the change of the spectrum (full frequency band) can be observed from the spectrum-amplitude dimension, and the change of the instantaneous frequency can be observed from the time-spectrum dimension.

[0003] In the prior art, signal analysis can be used for content analysis of image signals. For example, given the combination of image processing types that the currently acquired picture has gone through, that is, after multiple image processes, but if the order of these image processes is not determined, it is still very difficult for the monitoring end to clarify the true content of the currently acquired picture. However, there is a lack of an effective signal analysis mechanism for determining the order of multiple image processes in the prior art. Summary of the Invention

[0004] To solve the technical problems in related fields, the present invention provides a real-time generated picture information extraction system. By parallelly inputting the content data of each copy of the standard original image of the timing acquisition mechanism, the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points into the artificial intelligence model, the combination of image processing types passed by the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture is obtained. The combination of image processing types is represented by sequentially connecting multiple type names corresponding to multiple image processing types, so as to perform intelligent analysis on the image processing types and the image processing sequence passed by the real-time generated picture, and help the monitoring end clarify the real content of the real-time generated picture. Among them, the standard original image of the timing acquisition mechanism is the RAW image including only the standard graphics output by the sensor device of the timing acquisition mechanism, and the content data of each copy of the standard original image of the timing acquisition mechanism are the curvature values at each pixel point evenly distributed on the contour of the standard graphics in the standard original image of the timing acquisition mechanism, as well as the contrast and signal-to-noise ratio of the standard original image of the timing acquisition mechanism, thereby providing key basic information for the intelligent analysis of the combination of image processing types, and customizing the artificial intelligence model structure for performing the intelligent analysis of the combination of image processing types. Specifically, the artificial intelligence model is a deep neural network after each learning process, and the number of learning processes of the deep neural network is proportional to the number of pixel columns of the real-time generated picture, so as to design artificial intelligence models with different structures for different timing acquisition mechanisms.

[0005] According to the present invention, a real-time generated picture information extraction system is provided. The system includes:

[0006] A timing acquisition mechanism, which is arranged inside the school to collect pictures inside the school during the school time segments, and is used to obtain and output corresponding real-time generated pictures;

[0007] A first parsing mechanism, which is connected to the timing acquisition mechanism and is used to obtain the content data of each copy of the standard original image of the timing acquisition mechanism. Among them, the standard original image of the timing acquisition mechanism is the RAW image including only the standard graphics output by the sensor device of the timing acquisition mechanism;

[0008] A second parsing mechanism, which is connected to the timing acquisition mechanism and is used to obtain the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points;

[0009] An information conversion device, which is respectively connected to the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism, is configured to parallelly input the content data of each part of the standard original image of the timing acquisition mechanism, the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points into an artificial intelligence model, so as to obtain the combination of image processing types passed from the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture;

[0010] Among them, the combination of image processing types passed from the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture includes: the combination of image processing types is represented by sequentially connecting multiple type names corresponding to multiple image processing types;

[0011] Among them, the combination of image processing types is represented by sequentially connecting multiple type names corresponding to multiple image processing types includes: the image processing type is an image filtering type, and multiple image processing types are several processing types among median filtering processing, alpha mean filtering processing, adaptive filtering processing, bilateral filtering processing, guided filtering processing, and box filtering processing;

[0012] Among them, obtaining the content data of each part of the standard original image of the timing acquisition mechanism, where the standard original image of the timing acquisition mechanism is a RAW image output by the sensor device of the timing acquisition mechanism that only includes a standard graphic includes: the content data of each part of the standard original image of the timing acquisition mechanism is the curvature values at each pixel point evenly distributed on the contour of the standard graphic in the standard original image of the timing acquisition mechanism, as well as the contrast and signal-to-noise ratio of the standard original image of the timing acquisition mechanism.

[0013] Thus, the present invention at least has the following main inventive concepts:

[0014] The first point: Parallelly input the content data of each part of the standard original image of the timing acquisition mechanism, the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points into an artificial intelligence model, so as to obtain the combination of image processing types passed from the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture, and the combination of image processing types is represented by sequentially connecting multiple type names corresponding to multiple image processing types, thereby performing intelligent analysis on the image processing types and the image processing sequence passed by the real-time generated picture to help the monitoring end clarify the real content of the real-time generated picture;

[0015] Second: Specifically, the standard original image of the timing acquisition mechanism is the RAW image that only includes the standard graphics output by the sensor device of the timing acquisition mechanism, and each content data of the standard original image of the timing acquisition mechanism is the curvature value at each pixel point evenly distributed on the contour of the standard graphics in the standard original image of the timing acquisition mechanism, as well as the contrast and signal-to-noise ratio of the standard original image of the timing acquisition mechanism, thereby providing key basic information for the intelligent analysis of the image processing type combination;

[0016] Third: Customize the artificial intelligence model structure for performing intelligent analysis of the image processing type combination. Specifically, the artificial intelligence model is a deep neural network after each learning process, and the number of learning processes of the deep neural network is proportional to the number of pixel columns of the real-time generated image, so as to design artificial intelligence models with different structures for different timing acquisition mechanisms. Specific Embodiment

[0017] Hereinafter, embodiments of the real-time generated picture information extraction system of the present invention will be described in detail.

[0018] The real-time generated picture information extraction system shown according to the primary embodiment of the present invention includes:

[0019] A timing acquisition mechanism, which is arranged inside the school to collect pictures inside the school during the school time segments, and is used to obtain and output corresponding real-time generated pictures;

[0020] Specifically, the timing acquisition mechanism, which is arranged inside the school to collect pictures inside the school during the school time segments, and is used to obtain and output corresponding real-time generated pictures includes: the timing acquisition mechanism is internally provided with a timing component, a clock generation component, and a photoelectric induction component;

[0021] A first parsing mechanism, which is connected to the timing acquisition mechanism and is used to obtain each content data of the standard original image of the timing acquisition mechanism. Among them, the standard original image of the timing acquisition mechanism is the RAW image that only includes the standard graphics output by the sensor device of the timing acquisition mechanism;

[0022] A second parsing mechanism, which is connected to the timing acquisition mechanism and is used to obtain the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature value at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points;

[0023] An information conversion device, which is respectively connected to the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism, is configured to parallelly input each piece of content data of the standard original image of the timing acquisition mechanism, the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points into an artificial intelligence model, so as to obtain the combination of image processing types passed from the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture;

[0024] Among them, the combination of image processing types passed from the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture includes: the combination of image processing types is represented by sequentially connecting multiple type names corresponding to multiple image processing types;

[0025] Among them, the combination of image processing types is represented by sequentially connecting multiple type names corresponding to multiple image processing types, including: the image processing type is an image filtering type, and multiple image processing types are several processing types among midpoint filtering processing, alpha mean filtering processing, adaptive filtering processing, bilateral filtering processing, guided filtering processing, and box filtering processing;

[0026] Among them, obtaining each piece of content data of the standard original image of the timing acquisition mechanism, where the standard original image of the timing acquisition mechanism is a RAW image output by the sensor device of the timing acquisition mechanism and only includes standard graphics, includes: each piece of content data of the standard original image of the timing acquisition mechanism is the curvature values at each pixel point evenly distributed on the contour of the standard graphics in the standard original image of the timing acquisition mechanism, as well as the contrast and signal-to-noise ratio of the standard original image of the timing acquisition mechanism;

[0027] Among them, parallelly inputting each piece of content data of the standard original image of the timing acquisition mechanism, the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points into an artificial intelligence model, so as to obtain the combination of image processing types passed from the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture includes: the artificial intelligence model is a deep neural network after each learning process, and the number of learning processes of the deep neural network is proportional to the number of pixel columns of the real-time generated picture.

[0028] Compared with the primary embodiment of the present invention, the real-time generated picture information extraction system in the secondary embodiment of the present invention may further include:

[0029] A synchronous drive device, which is arranged near the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism and is respectively connected to the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism;

[0030] Among them, the synchronous drive device, which is arranged near the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism and is respectively connected to the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism, includes: the synchronous drive device is used to respectively realize the synchronous drive control of two-by-two devices of the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism.

[0031] Compared with the primary embodiment of the present invention, the real-time generated picture information extraction system in the secondary embodiment of the present invention may further include:

[0032] A circuit supply device, which is arranged near the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism and is respectively connected to the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism;

[0033] Among them, the circuit supply device, which is arranged near the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism and is respectively connected to the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism, includes: the circuit supply device is used to respectively provide the working voltage values required by the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism.

[0034] Next, the specific structure of the real-time generated picture information extraction system of the present invention will be further described.

[0035] In the real-time generated picture information extraction system according to various embodiments of the present invention:

[0036] An ASIC chip is used to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism to obtain the output processing data respectively corresponding to the information conversion device, the timing acquisition mechanism, the first analysis mechanism and the second analysis mechanism.

[0037] In the real-time generated picture information extraction system according to various embodiments of the present invention:

[0038] Using an ASIC chip to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively includes: performing maximum value filtering processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.

[0039] In the real-time generated picture information extraction system according to various embodiments of the present invention:

[0040] Using an ASIC chip to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively includes: performing minimum value filtering processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.

[0041] In the real-time generated picture information extraction system according to various embodiments of the present invention:

[0042] Using an ASIC chip to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively includes: performing median value filtering processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.

[0043] And in the real-time generated picture information extraction system according to various embodiments of the present invention:

[0044] An ASIC chip is used to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively, including: performing edge sharpening processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.

[0045] In addition, in the real-time generated picture information extraction system, the content data of each copy of the standard original image of the timing acquisition mechanism, the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly spaced on the contour of the target with the largest number of occupied pixel points are input into the artificial intelligence model in parallel, so as to obtain the combination of image processing types passed by the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture, including: using an information processing formula to represent the information corresponding relationship that the number of learning processes of the deep neural network is proportional to the number of pixel columns of the real-time generated picture.

[0046] By using the real-time generated picture information extraction system of the present invention, aiming at the technical problem in the prior art that there is a lack of an accurate identification mechanism for the respective processing sequences of multiple image processing types passed by the currently acquired image, the combination of image processing types passed by the RAW image corresponding to the real-time generated picture is intelligently identified by using a specifically designed artificial intelligence model based on customized and screened visual data. The image processing type combination is represented by sequentially connecting multiple type names corresponding to multiple image processing types, so as to help the monitoring end clarify the real content of the real-time generated picture, and solve the above technical problem.

[0047] Exemplary embodiments of the present invention have been described. It should also be noted that it is obvious to those skilled in the art that various improvements can be made without departing from the spirit and scope of the present invention defined by the boundaries and limits of the appended claims.

Claims

1. A real-time generated picture information extraction system, characterized in that, The system includes: A timing acquisition mechanism, which is set inside the school to collect pictures inside the school during segmented school hours, and is used to obtain and output corresponding real-time generated pictures; A first parsing mechanism, connected to the timing acquisition mechanism, and is used to obtain each content data of the standard original image of the timing acquisition mechanism. Among them, the standard original image of the timing acquisition mechanism is a RAW image output by the sensor device of the timing acquisition mechanism that only includes standard graphics; A second parsing mechanism, connected to the timing acquisition mechanism, and is used to obtain the maximum gray value, minimum gray value, background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points; An information conversion device, respectively connected to the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism, and is used to parallelly input each content data of the standard original image of the timing acquisition mechanism, the maximum gray value, minimum gray value, background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly distributed on the contour of the target with the largest number of occupied pixel points into the artificial intelligence model, so as to obtain the combination of image processing types passed from the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture; Among them, the combination of image processing types passed from the RAW image corresponding to the real-time generated picture output by the artificial intelligence model to the real-time generated picture includes: the combination of image processing types is represented by sequentially connecting multiple type names corresponding to multiple image processing types; Among them, the combination of image processing types is represented by sequentially connecting multiple type names corresponding to multiple image processing types includes: the image processing type is an image filtering type, and the multiple image processing types are several processing types among median filtering processing, alpha mean filtering processing, adaptive filtering processing, bilateral filtering processing, guided filtering processing, and box filtering processing; Among them, obtaining each content data of the standard original image of the timing acquisition mechanism, where the standard original image of the timing acquisition mechanism is a RAW image output by the sensor device of the timing acquisition mechanism that only includes standard graphics includes: each content data of the standard original image of the timing acquisition mechanism is the curvature values at each pixel point evenly distributed on the contour of the standard graphics in the standard original image of the timing acquisition mechanism, as well as the contrast and signal-to-noise ratio of the standard original image of the timing acquisition mechanism; Parallelly input the content data of each copy of the standard original image of the timing acquisition mechanism, the maximum gray value, the minimum gray value, the background area ratio of the real-time generated picture, and the curvature values at each pixel point evenly spaced on the contour of the target with the largest number of occupied pixel points into the artificial intelligence model to obtain the image processing type combination passed by the RAW image corresponding to the real-time generated picture to the real-time generated picture output by the artificial intelligence model. The artificial intelligence model is a deep neural network after each learning process, and the number of learning processes of the deep neural network is proportional to the number of pixel columns of the real-time generated picture; Use an information processing formula to represent the information correspondence relationship in which the number of learning processes of the deep neural network is proportional to the number of pixel columns of the real-time generated picture.

2. The real-time generated picture information extraction system according to claim 1, wherein The system further includes: Synchronous driving devices, arranged near the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism and respectively connected to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism; Among them, the synchronous driving devices, arranged near the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism and respectively connected to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism include: The synchronous driving devices are used to respectively realize the synchronous driving control of two-by-two devices of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism.

3. The real-time generated picture information extraction system according to claim 1, wherein The system further includes: Circuit supply devices, arranged near the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism and respectively connected to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism; Among them, the circuit supply devices, arranged near the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism and respectively connected to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism include: The circuit supply devices are used to respectively provide the working voltage values required by the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism.

4. The real-time generated picture information extraction system according to any one of claims 1-3, characterized in that: Use an ASIC chip to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.

5. The real-time generated picture information extraction system according to claim 4, characterized in that: Using an ASIC chip to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively includes: performing maximum value filtering processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.

6. The real-time generated picture information extraction system according to claim 4, wherein: Using an ASIC chip to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively includes: performing minimum value filtering processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.

7. The real-time generated picture information extraction system according to claim 4, wherein: Using an ASIC chip to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively includes: performing median filtering processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.

8. The real-time generated picture information extraction system according to claim 4, wherein: Using an ASIC chip to perform image data processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively includes: performing edge sharpening processing on the output data of the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism to obtain the output processing data corresponding to the information conversion device, the timing acquisition mechanism, the first parsing mechanism, and the second parsing mechanism respectively.