Image restoration method based on illumination correction

Abstract

The invention provides an image restoration method based on illumination correction, belonging to the technical field of image processing. The image restoration method comprises the following steps: 1, carrying out Lee filtration on degraded images to filter the influence of noise; 2, adopting an image subblock method to determine the membership degree mu (xi, theta) of an illumination distribution parameter theta, taking the membership degree as the probability of occurrence of input values and obtaining the probability when all subblocks occur synchronously; 3, carrying out integration on the degraded images after filtration to obtain entropy of an illumination distribution matrix of the degraded images; 4, estimating the value of the illumination distribution parameter theta according to the maximum-minimum mu / E principle; and 5, utilizing the estimated illumination distribution parameter to obtain a non-ideal illumination distribution image which is the same as the original image in size, and then subtracting the non-ideal illumination distribution image from an observed image, namely a real image. The image restoration method has the advantage of good restoration capability for edge details and smooth areas, is obviously superior to the common image restoration method on restoration time and restoration effect, and can be used for restoration of space images, aerial images, underwater or medical images and common photographic images.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to an image restoration method based on illumination correction. Background technique [0002] Because the image is affected by unideal factors in the optical imaging process, the quality of the obtained image is degraded. In order to restore the degraded image information, it is necessary to use image restoration technology to improve the quality of the image. When imaging different targets under the same lighting conditions, different images usually have similar degraded features. If the information of these similar degraded features can be fully utilized for illumination correction, it will have a great impact on the image restoration results. improve. The more representative illumination correction methods include bicubic interpolation method, homomorphic filtering method and top-hat transformation method, etc. The bicubic interpolation method processes th...

AI Technical Summary

Problems solved by technology

The bicubic interpolation method processes the illumination model of the image into an additive form, uses the bicubic surface model to regress the illumination distribution, and realizes image restoration by subtracting the estimated illumination from the original image. This method is effective for the restoration of most images. , but the calculation of bicubic interpolation is large, so it is not widely used; the homomorphic filtering method processes the illumination model of the image into a multiplicative form, performs logarithmic transformation on the illumination model, and then performs high-pass filtering on the image to achieve image restoration. For many imaging results where the illumination distribution cannot satisfy the multiplication model, the restoration effect of this method is not ideal; the top-hat transformation method uses the principle of morphology to estimate the illumination distribution model through the top-hat transformation, which requires a large amount of calculation and cannot for real-time processing

Images