Image data intelligent processing selection system

Abstract

The invention relates to an image data intelligent processing selection system, which comprises a traversal selection mechanism, the sharpening algorithm is used for sharpening a current sharpening algorithm according to a binary numerical value of an algorithm code corresponding to the current sharpening algorithm, color channel numerical values corresponding to pixel points of a signal restoration image, and the maximum noise amplitude of the signal restoration image. Intelligently analyzing the contrast ratio of a sharpened image obtained after the sharpening operation is performed on the signal restoration image by adopting the current sharpening algorithm according to the minimum noise amplitude and the number of pixel points occupied by the background area; and the parameter display mechanism is used for performing associated display on the contrast of each sharpened image corresponding to each sharpening algorithm and each sharpening algorithm. According to the image data intelligent processing and selecting system, the contrast ratio of each sharpened image corresponding to each sharpening algorithm and each sharpening algorithm can be displayed in an associated manner, so that the sharpening algorithm with the best sharpening effect is provided for the current image data.

Description

Technical Field

[0001] The present invention relates to the field of image sharpening, and in particular to an image data intelligent processing and selection system. Background Art

[0002] Image sharpening is to compensate for the contour of the image, enhance the edge of the image and the grayscale jump part, so that the image becomes clear. It is divided into two categories: spatial domain processing and frequency domain processing. Image sharpening is to highlight the edge, contour, or characteristics of certain linear target elements on the image. This filtering method increases the contrast between the edge of the object and the surrounding pixels, so it is also called edge enhancement.

[0003] The existing image sharpening algorithms include: 1. An image sharpening algorithm based on HPF, disclosed in patent document CN107784641A, and its process is as follows: (1) Preprocess the image; (2) Perform HPF convolution and processing on the image; (3) Determine the Conv convolution image parameters TH, G, SHF, MAX. (4) Superimpose the image. The advantage is that the image grayscale information is used to adaptively remove noise and enhance the edge of the image after HPF convolution, and the original image is superimposed with the convolution image to obtain the sharpened image. The calculation method of the algorithm adopts full integer calculation, which speeds up the algorithm's operation speed. At the same time, it also adapts to some platforms that can only perform integer operations, improves the versatility of the algorithm, and enables the algorithm to have better portability and faster operation speed. 2. The image boundary sharpening algorithm, device and control system based on FPGA language are disclosed in the patent document CN116523764A. By acquiring the picture, the RGB signal is converted into YUV data; based on the Laplace operator template, the Y component of the current pixel in the picture is convolved to calculate and obtain the MaxLine value of the current pixel; the MaxLine values ​​of all other pixels within N×N of the current pixel are obtained, and the sum of the MaxLine values ​​of all other pixels is calculated; if the sum of all other pixels is greater than 0 and the MaxLine value of the current pixel is equal to 1, the current pixel is identified as a boundary point that needs to be sharpened, and the final sharpening value is calculated. The boundary points that meet the sharpening conditions are calculated and found through the Laplace operator template and sharpened, which is used for coherent sharpening to achieve a higher image effect. The advantage is that the algorithm occupies less resources and can reduce the noise of isolated points.

[0004] However, in actual operation, different image contents of the same type may need to choose the best sharpening algorithm suitable for themselves to ensure the best sharpening effect of different image contents and improve the image quality of different image contents. Using the same sharpening mode to perform the same sharpening processing on all image contents may result in unsatisfactory results. How to choose the best sharpening algorithm suitable for different image contents of the same type is one of the difficult problems that need to be solved in the existing technology. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides an image data intelligent processing selection system, which can associate the contrast of each sharpening algorithm with the various sharpening algorithms for display operations, wherein, for each sharpening algorithm, according to the binary value of the corresponding algorithm code, the color channel values ​​corresponding to each pixel of the signal restoration image, the maximum noise amplitude of the signal restoration image, the minimum noise amplitude and the number of pixels occupied by the background area, the contrast of the image after the sharpening operation obtained by performing the sharpening operation on the signal restoration image is intelligently analyzed, thereby providing the sharpening algorithm with the best sharpening effect for the current image data.

[0006] The system comprises: A content receiving device, used for selecting a signal to restore an image to be processed based on a user's operation; A data interpolation device, connected to the content receiving device, for performing nearest neighbor interpolation processing on the received signal restoration image to obtain and output a corresponding nearest neighbor interpolation image; A content sharpening device, connected to the data interpolation device, for performing a sharpening process based on a Kirsch operator on the received nearest neighbor interpolation image to obtain and output a corresponding content sharpened image; A point image restoration device, connected to the content sharpening device, for performing single or multiple point image restoration processing on the received content sharpened image to obtain and output a corresponding signal restoration image; A traversal selection mechanism is connected to the point image restoration device and includes a channel analysis unit, a noise analysis unit, a model building unit and a sharpening prediction unit. The sharpening prediction unit is respectively connected to the channel analysis unit, the noise analysis unit and the model building unit. The traversal selection mechanism is used to determine the color channel values ​​corresponding to each pixel of the signal restoration image, analyze the maximum noise amplitude, the minimum noise amplitude and the number of pixels occupied by the background area of ​​the signal restoration image, and intelligently analyze the contrast of the image after the sharpening operation obtained by using the current sharpening algorithm to perform a sharpening operation on the signal restoration image according to the binary value of the algorithm code corresponding to the current sharpening algorithm, the color channel values ​​corresponding to each pixel of the signal restoration image, the maximum noise amplitude, the minimum noise amplitude and the number of pixels occupied by the background area of ​​the signal restoration image, wherein the intelligent analysis adopts a feedforward neural network model, and the number of training times of the feedforward neural network model is positively correlated with the maximum noise amplitude of the signal restoration image; A parameter display mechanism, connected to the traversal selection mechanism, for displaying the contrast of each sharpening image corresponding to each sharpening algorithm in association with the sharpening algorithms; Among them, the operation of associating the contrast of each sharpening algorithm with the various sharpening algorithms for display also includes: for each sharpening algorithm, based on the binary value of the algorithm code corresponding to it, the color channel values ​​corresponding to each pixel point of the signal restoration image, the maximum noise amplitude of the signal restoration image, the minimum noise amplitude and the number of pixels occupied by the background area, intelligently analyzing the contrast of the image after the sharpening operation obtained by performing the sharpening operation on the signal restoration image.

[0007] The image data intelligent processing selection system of the present invention is intelligently designed and widely used. Since the contrast of each sharpening operation image corresponding to each sharpening algorithm can be associated with various sharpening algorithms and displayed, the sharpening algorithm with the best sharpening effect is provided for the current image data. DETAILED DESCRIPTION

[0008] The following is a detailed description of an embodiment of the image data intelligent processing and selection system of the present invention.

[0009] A. Embodiment

[0010] According to embodiment A of the present invention, an image data intelligent processing and selection system is shown, the system comprising: A content receiving device, used for selecting a signal to restore an image to be processed based on a user's operation; A data interpolation device, connected to the content receiving device, for performing nearest neighbor interpolation processing on the received signal restoration image to obtain and output a corresponding nearest neighbor interpolation image; For example, a CPLD device may be selected to implement the data interpolation device, connected to the content receiving device, and used to perform nearest neighbor interpolation processing on the received signal restoration image to obtain and output a corresponding nearest neighbor interpolation image; A content sharpening device, connected to the data interpolation device, for performing a sharpening process based on a Kirsch operator on the received nearest neighbor interpolation image to obtain and output a corresponding content sharpened image; A point image restoration device, connected to the content sharpening device, for performing single or multiple point image restoration processing on the received content sharpened image to obtain and output a corresponding signal restoration image; A traversal selection mechanism is connected to the point image restoration device and includes a channel analysis unit, a noise analysis unit, a model building unit and a sharpening prediction unit. The sharpening prediction unit is respectively connected to the channel analysis unit, the noise analysis unit and the model building unit. The traversal selection mechanism is used to determine the color channel values ​​corresponding to each pixel of the signal restoration image, analyze the maximum noise amplitude, the minimum noise amplitude and the number of pixels occupied by the background area of ​​the signal restoration image, and intelligently analyze the contrast of the image after the sharpening operation obtained by using the current sharpening algorithm to perform a sharpening operation on the signal restoration image according to the binary value of the algorithm code corresponding to the current sharpening algorithm, the color channel values ​​corresponding to each pixel of the signal restoration image, the maximum noise amplitude, the minimum noise amplitude and the number of pixels occupied by the background area of ​​the signal restoration image, wherein the intelligent analysis adopts a feedforward neural network model, and the number of training times of the feedforward neural network model is positively correlated with the maximum noise amplitude of the signal restoration image; A parameter display mechanism, connected to the traversal selection mechanism, for displaying the contrast of each sharpening image corresponding to each sharpening algorithm in association with the sharpening algorithms; The operation of associating the contrast ratios of the images after sharpening operations corresponding to the various sharpening algorithms with the various sharpening algorithms for display also includes: for each sharpening algorithm, intelligently analyzing the contrast ratio of the image after sharpening operations obtained by performing the sharpening operation on the signal restoration image according to the binary value of the algorithm code corresponding to the algorithm, the color channel values ​​corresponding to the pixels of the signal restoration image, the maximum noise amplitude and the minimum noise amplitude of the signal restoration image, and the number of pixels occupied by the background area; Among them, the contrast of each sharpening operation image corresponding to each sharpening algorithm is associated with various sharpening algorithms for display operation, including: different sharpening algorithms are respectively displayed corresponding to the contrast of different sharpening operations of the image, and each sharpening algorithm is followed by the contrast of the image after the sharpening operation corresponding to it.

[0011] B. Embodiment

[0012] According to Embodiment B of the present invention, an image data intelligent processing and selection system is shown, which may also include the following components: a current detection device, disposed near the channel analysis unit, the noise analysis unit, the model building unit, and the sharpening prediction unit and connected to the channel analysis unit, the noise analysis unit, the model building unit, and the sharpening prediction unit, respectively; Among them, the current detection device is arranged near the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit and is respectively connected to the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit, including: the current detection device is used to respectively realize the on-site measurement of the current current of the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit.

[0013] C Example

[0014] According to the C embodiment of the present invention, an image data intelligent processing and selection system is shown, which may also include the following components: a bandwidth detection device, which is arranged near the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit and is connected to the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit respectively; Among them, the bandwidth detection device is arranged near the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit and is respectively connected to the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit, including: the bandwidth detection device is used to respectively realize on-site measurement of the current bandwidth of the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit.

[0015] Next, the specific structure of the image data intelligent processing and selection system of the present invention will be further described.

[0016] In the image data intelligent processing selection system according to various embodiments of the present invention: A field programmable gate array is used to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

[0017] In the image data intelligent processing selection system according to various embodiments of the present invention: Using a field programmable gate array to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively includes: performing statistical sorting filtering on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

[0018] In the image data intelligent processing selection system according to various embodiments of the present invention: Using a field programmable gate array to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively includes: performing gradient sharpening filtering on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

[0019] In the image data intelligent processing selection system according to various embodiments of the present invention: Using a field programmable gate array to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively includes: performing Butterworth low-pass filtering on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

[0020] And in the image data intelligent processing selection system according to various embodiments of the present invention: Using a field programmable gate array to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively includes: performing Gaussian high-pass filtering on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

[0021] In addition, in the image data intelligent processing and selection system, various sharpening algorithms include statistical sorting sharpening algorithm, maximum sharpening algorithm, minimum sharpening algorithm, midpoint sharpening algorithm, alpha mean sharpening algorithm, adaptive sharpening algorithm, bilateral sharpening algorithm, guided sharpening algorithm and box sharpening algorithm.

[0022] The technical effects of the present invention are: (1) performing an associated display operation on the contrast ratios of the images after sharpening operations corresponding to the various sharpening algorithms and the various sharpening algorithms, wherein, for each sharpening algorithm, intelligently analyzing the contrast ratios of the images after sharpening operations obtained by performing the sharpening operations on the signal restoration image based on the binary values ​​of the algorithm codes corresponding to the algorithms, the values ​​of the color channels corresponding to the pixels of the signal restoration image, the maximum noise amplitude and the minimum noise amplitude of the signal restoration image, and the number of pixels occupied by the background area, thereby providing a sharpening algorithm with the best sharpening effect for the current image data; (2) introducing a traversal selection mechanism including a channel analysis unit, a noise analysis unit, a model building unit and a sharpening prediction unit, wherein the sharpening prediction unit is connected to the channel analysis unit, the noise analysis unit and the model building unit respectively, thereby providing hardware resources for the advance prediction of the sharpening effect of the sharpening algorithm; (3) Determine the values ​​of each color channel corresponding to each pixel of the signal restoration image, analyze the maximum noise amplitude, the minimum noise amplitude and the number of pixels occupied by the background area of ​​the signal restoration image, and intelligently analyze the contrast of the image after the sharpening operation is performed on the signal restoration image using the current sharpening algorithm according to the binary value of the algorithm code corresponding to the current sharpening algorithm, the values ​​of each color channel corresponding to each pixel of the signal restoration image, the maximum noise amplitude, the minimum noise amplitude and the number of pixels occupied by the background area of ​​the signal restoration image; (4) The intelligent analysis adopts a feedforward neural network model, and the number of training times of the feedforward neural network model is positively correlated with the maximum noise amplitude of the signal restoration image, thereby realizing a customized design of the structure of the feedforward neural network model.

[0023] The present embodiments are to be considered in all respects as illustrative and not restrictive, and all changes coming within the meaning and equivalency range of the claims are intended to be embraced within their scope.

Claims

1. An image data intelligent processing and selection system, characterized in that: The system comprises: A content receiving device, used for selecting a signal to restore an image to be processed based on a user's operation; A data interpolation device, connected to the content receiving device, for performing nearest neighbor interpolation processing on the received signal restoration image to obtain and output a corresponding nearest neighbor interpolation image; A content sharpening device, connected to the data interpolation device, for performing a sharpening process based on a Kirsch operator on the received nearest neighbor interpolation image to obtain and output a corresponding content sharpened image; A point image restoration device, connected to the content sharpening device, for performing single or multiple point image restoration processing on the received content sharpened image to obtain and output a corresponding signal restoration image; A traversal selection mechanism is connected to the point image restoration device and includes a channel analysis unit, a noise analysis unit, a model building unit and a sharpening prediction unit. The sharpening prediction unit is respectively connected to the channel analysis unit, the noise analysis unit and the model building unit. The traversal selection mechanism is used to determine the color channel values ​​corresponding to each pixel of the signal restoration image, analyze the maximum noise amplitude, the minimum noise amplitude and the number of pixels occupied by the background area of ​​the signal restoration image, and intelligently analyze the contrast of the image after the sharpening operation obtained by using the current sharpening algorithm to perform a sharpening operation on the signal restoration image according to the binary value of the algorithm code corresponding to the current sharpening algorithm, the color channel values ​​corresponding to each pixel of the signal restoration image, the maximum noise amplitude, the minimum noise amplitude and the number of pixels occupied by the background area of ​​the signal restoration image, wherein the intelligent analysis adopts a feedforward neural network model, and the number of training times of the feedforward neural network model is positively correlated with the maximum noise amplitude of the signal restoration image; A parameter display mechanism, connected to the traversal selection mechanism, for displaying the contrast of each sharpening image corresponding to each sharpening algorithm in association with the sharpening algorithms; Among them, the operation of associating the contrast of each sharpening algorithm with the various sharpening algorithms for display also includes: for each sharpening algorithm, based on the binary value of the algorithm code corresponding to it, the color channel values ​​corresponding to each pixel point of the signal restoration image, the maximum noise amplitude of the signal restoration image, the minimum noise amplitude and the number of pixels occupied by the background area, intelligently analyzing the contrast of the image after the sharpening operation obtained by performing the sharpening operation on the signal restoration image.

2. The image data intelligent processing and selection system according to claim 1, characterized in that: The contrast of each sharpening operation image corresponding to each sharpening algorithm is associated with various sharpening algorithms for display operation, including: different sharpening algorithms are respectively displayed corresponding to the contrast of different sharpening operation images, and each sharpening algorithm is followed by the contrast of the image after the sharpening operation corresponding to it.

3. The image data intelligent processing and selection system according to claim 2, characterized in that: The system further comprises: a current detection device, disposed near the channel analysis unit, the noise analysis unit, the model building unit, and the sharpening prediction unit and connected to the channel analysis unit, the noise analysis unit, the model building unit, and the sharpening prediction unit, respectively; Among them, the current detection device is arranged near the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit and is respectively connected to the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit, including: the current detection device is used to respectively realize the on-site measurement of the current current of the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit.

4. The image data intelligent processing and selection system according to claim 2, characterized in that: The system further comprises: a bandwidth detection device, which is arranged near the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit and is connected to the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit respectively; Among them, the bandwidth detection device is arranged near the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit and is respectively connected to the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit, including: the bandwidth detection device is used to respectively realize on-site measurement of the current bandwidth of the channel analysis unit, the noise analysis unit, the model building unit and the sharpening prediction unit.

5. The image data intelligent processing and selection system according to any one of claims 2 to 4, characterized in that: A field programmable gate array is used to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

6. The image data intelligent processing and selection system according to claim 5, characterized in that: Using a field programmable gate array to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively includes: performing statistical sorting filtering on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

7. The image data intelligent processing and selection system according to claim 5, characterized in that: Using a field programmable gate array to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively includes: performing gradient sharpening filtering on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

8. The image data intelligent processing and selection system according to claim 5, characterized in that: Using a field programmable gate array to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively includes: performing Butterworth low-pass filtering on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.

9. The image data intelligent processing and selection system according to claim 5, characterized in that: Using a field programmable gate array to perform image data processing on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively includes: performing Gaussian high-pass filtering on the output data of the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit to obtain the output processing data corresponding to the channel parsing unit, the noise parsing unit, the model building unit and the sharpening prediction unit respectively.