为解决矿井下复杂光照条件导致视频监控系统中的图像降质问题，提出基于自适应小波变换的煤矿降质图像模糊增强算法。首先，通过多尺度小波分解将矿井降质图像分解为低频子图和不同尺度的高频子图，采用贝叶斯估计的小波收缩阈值方法，自适应调整不同尺度下高频子图的小波阈值;其次，设计了引入自适应权值因子和自适应增强系数的自适应小波阈值函数，在保持了阈值函数连续性的同时也能够避免产生固定偏差，从而实现了对不同尺度下高频子图的收缩阈值滤波和非线性增强;接着，采用双边滤波算法估计并去除低频子图中的照度分量，并对处理后的低频子图和各尺度高频子图进行小波重构，获取增强后的小波重构图像;最后，采用改进的隶属度函数和模糊增强算子对小波重构图像的亮度分量进行调整，得到最终增强图像。采用主观视觉和客观评价指标对降质图像增强实验进行分析验证，实验结果表明:所述算法具有最优的图像增强效果，能够有效抑制图像噪声、增强图像细节特征、降低图像失真、改善降质图像视觉效果，且克服了传统图像增强算法在矿井复杂光照条件下的局限性，具有较强的鲁棒性。其综合性能指标较CLAHE，SSR，MSR，BF-DCP，DGR，MSWT及PGCHE七种算法，分别提高4.42%，4.95%，15.35%，196.60%，88.93%，10.52%和12.10%。

In order to solve the problem of image degradation in video monitoring systems caused by complex lighting conditions in mines,a fuzzy enhancement algorithm for coal mine degradation images based on adaptive wavelet transform is proposed.Firstly,the degraded image is decomposed into low-frequency sub-graph and high-frequency sub-graphs of different scales by multi-scale wavelet decomposition,and the wavelet shrinkage threshold method of Bayesian estimation is used to adaptively adjust the wavelet threshold at different scales.Secondly,an adaptive wavelet threshold function that introduces adaptive weight factor and adaptive enhancement coefficient is designed,which not only maintains the continuity of the threshold function and avoids fixed deviations,but also implements contraction threshold filtering and non-linear en-hancement for high-frequency sub-graphs of different scales.Thirdly,a bilateral filtering algorithm is used to estimate and remove the illuminance component in the low-frequency sub-graph,and the wavelet reconstruction is performed on the processed low-frequency sub-graph and high-frequency sub-graphs of each scale to obtain enhanced wavelet reconstruction images.Finally,the improved membership function and fuzzy enhancement operator are used to adjust the brightness component of wavelet reconstructed image to obtain the final enhanced image.The subjective vision and objective evaluation indicators are used to analyze the results of degraded image enhancement experiments.The proposed algorithm has the best image enhancement effect,which can effectively suppress image noise,enhance detailed information,reduce image distortion,improve the visual effect of degraded images,overcome the limitations of traditional image enhancement algorithms under complex lighting conditions in mines,and has strong robustness.Compared with CLAHE,SSR,MSR,BF-DCP,DGR,MSWT and PGCHE,its comprehensive performance indicators have been improved by 4.42%,4.95%,15.35%,196.60%,88.93%,10.52% and 12.10% respectively.