【摘要】 针对煤矿主皮带图像的低光照、多尘雾、噪声大等问题,提出一种基于加权引导滤波的二维变分模态算法对主皮带图像进行增强。该算法对预处理的图像进行二维自适应非递归变分分解,对分解后的低频子模态进行加权引导滤波,增强图像的边缘细节以提高图像清晰度;并采用去噪能力强维纳斯滤波器。通过与自适应双边滤波和加权引导滤波技术仿真对比,该方法在图像边缘细节、滤除噪声等方面视觉效果不错,同时客观的峰值信噪比参数较高,均方差参数较低。更多还原

【Abstract】 A two-dimensional variational modal algorithm based on weighted guidance filtering is proposed to enhance the main belt image for low light,dust and fog of the main belt image of coal mine.The algorithm performs two-dimensional adaptive non-recursive variational decomposition on the preprocessed image,performs weighted guided filtering on the decomposed low-frequency sub-modality,enhances the edge detail of the image to improve image sharpness,and adopts filter with strong denoising ability.Compared with the adaptive bilateral filtering and weighted guided filtering technology,the method has good visual effects in image edge detail and filtering noise,the objective peak signal-to-noise ratio parameter is higher,and the mean square error parameter is lower.更多还原