Research On Denoising And Enhancement Algorithm Of Mine Image

At present,coal mining technology is in the process of transformation and upgrading,from dangerous manual production to safe production,and from mechanized and automatic mining to unmanned mining.In this process,image processing technology will play an important role.Due to the interference of uneven illumination,dark environment and high dust in the underground environment,the mine image has the characteristics of high noise and low contrast,which will seriously hinder the intelligent detection technology with image processing technology as the core.Based on this background,this paper studies the technology of mine image noise removal and uneven illumination image enhancement.The main contents of this paper are as follows:(1)Aiming at the common noise of mine image,a weighted curvature filtering algorithm based on the image median gray similarity function is proposed.The variance of the median gray similarity function depends on the highest frequency subband coefficient of wavelet transform,which can better prevent the image from being over smooth and improve the ability of removing salt and pepper noise.Therefore,the median gray similarity function is used to filter the local Gaussian curvature and the local total curvature respectively.The local weighted Gaussian curvature projection operator and the local weighted total variational curvature projection operator are iterated respectively until the total gradient energy of the output image reaches the stop condition.Experiments show that the proposed algorithm can effectively reduce the Gaussian noise and salt and pepper noise contained in mine image.(2)In order to solve the problem of uneven illumination of mine image,a dark area and detail enhancement algorithm based on HSV image is proposed.Firstly,a multi-scale Retinex(MSR)model weighted by the minimum perceptible difference(JND)coefficient is proposed to enhance the value(V).At the same time,the Laplace pyramid is used to extract the detail information of V,and the value and detail information after enhancement are linearly weighted to obtain the increment of V Then,the enhanced value is used to adjust the saturation(S).Finally,the enhanced value,the adjusted saturation and the original hue(H)are converted to RGB space to get the final enhanced image.The experimental results show that the proposed method can enhance the details of dark areas and effectively suppress the excessive enhancement of bright areas.