Boundary point processed histogram balancing method

Abstract

The invention provides a column diagram balancing method for processing boundary point, which is the method for processing the edge image pixels but not all image pixels, thus effectively restraining the spreading of background noise and effectively maintaining certain grayscale level not to be merged. At the same time, the image strengthening effect is controlled by a method for adding a control factor so that the typical value of the image has no far more difference with the typical value of the original image.

Description

technical field [0001] The invention relates to a histogram equalization method, in particular to a histogram equalization method for processing boundary points. Background technique [0002] Histogram equalization is a commonly used image enhancement technology. Its advantage is that it can automatically enhance the contrast of the entire image, it expands the dynamic range of the image, and the resulting image density distribution becomes flat, but its specific enhancement effect is not easy to control. The result of the processing is always a global equalized histogram, and for some gray levels with a small gray frequency, they will be merged, resulting in a decrease in resolution and a visually obvious sense of rigidity; after the histogram is equalized, the It will increase the background noise, and the average value of the processed image is always near the median value of the gray scale range, which has nothing to do with the average gray scale of the original image. ...

AI Technical Summary

Problems solved by technology

[0002] Histogram equalization is a commonly used image enhancement technology. Its advantage is that it can automatically enhance the contrast of the entire image, it expands the dynamic range of the image, and the resulting image density distribution becomes flat, but its specific enhancement effect is not easy to control. The result of the processing is always a global equalized histogram, and for some gray levels with a small gray frequency, they will be merged, resulting in a decrease in resolution and a visually obvious sense of rigidity; after the histogram is equalized, the It will increase the background noise, and the mean value of the processed image is always near the median value of the gray scale range, which has nothing to do with the average gray scale of the original image

[0003] The Chinese patent application No. 99122874.X proposes an image enhancement device that maintains the brightness of the input image and improves the contrast. A brightness difference compensator is designed to compensate the output image when the brightness of the output image is lower than the input image. Solve the problem that the brightness has nothing to do with the gray level of the original image, that is, the gray level combination causes the resolution to drop, but it cannot reduce the background noise

[0004] The Chinese patent application number 97111448.X proposes a method and device for low-pass filtering the image first and then performing histogram equalization, and then subtracting the low-pass filtered signal from the input image signal to superimpose the contrast-enhanced signal to solve the problem of large background noise, but it cannot solve the problem of gray level merging

Images