Enchantment of Ore-rock Fissure Images Based on Human Visual Perception and Improved Multi-scale Retinex Algorithm
Accurate and effective detection of ore-rock fissure plays an important role in the stability analysis of rock mass.Aiming at the problem of insufficient feature extraction in low resolution images acquired under low light conditions by deep network,a new algorithm for ore-rock fissure image enhancement based on human vision and improved multi-scale Retinex is proposed.Firstly,the low-resolution images of ore-rock fissure are preprocessed by combining human visual features and wavelet transform algorithm.Then,the improved multi-scale Retinex algorithm is used to estimate the noise from illumination and reflection components,and eliminate the interference to the image.Finally,CBAM(Convolutional Block Attention Module)is used to enhance the expression ability of denoised images in channel and spatial dimensions,and to achieve low-resolution image enhancement of mine and rock fissure images.The experimental results show that compared with the traditional Retinex algorithm,the improved multi-scale Retinex model can better extract image details,enhance image contrast and make cracks more clearly visible while maintaining the natural perception.
NETL
NSTL
万方数据
