Solid waste surface feature enhancement method based on gamma correction and wavelet transform

Abstract

The solid waste images obtained by the machine vision recognition system have many problems, such as uneven illumination, low contrast, and not apparent surface features, seriously affecting the accuracy of surface feature recognition. A method of combining gamma correction and wavelet transform is proposed to enhance the details of images. Firstly, the original image is transformed into hue-saturation-intensity color space, and the guided filtering method is used to separate the illumination form intensity components. An improved two-dimensional gamma function is constructed to balance brightness and enhance contrast. The gamma correction parameter is dynamically adjusted according to the distribution of illumination. Secondly, the wavelet transform is used to convert the corrected image into the frequency domain. An improved wavelet threshold algorithm is proposed to remove the noise introduced by gamma correction. Lastly, the wavelet transform is used to restructure intensity and convert the image to initial color space. Experimental results demonstrate that the proposed method could enhance the details of images under several different illumination conditions.