针对磁瓦表面缺陷图像整体光照不均匀、对比度较低、纹理细节信息不清晰等问题,提出了一种改进的同态滤波与连续均值量化变换(successive mean quantization transform,SMQT)相结合的磁瓦表面缺陷图像增强算法。首先,使用改进的巴特沃...针对磁瓦表面缺陷图像整体光照不均匀、对比度较低、纹理细节信息不清晰等问题,提出了一种改进的同态滤波与连续均值量化变换(successive mean quantization transform,SMQT)相结合的磁瓦表面缺陷图像增强算法。首先,使用改进的巴特沃斯传递函数替换传统同态滤波中的传递函数,使其将图像分解成高频图像和低频图像;然后,使用SMQT算法非线性拉伸高频图像的灰度范围使其扩展到整个灰度级区域,从而达到对高频图像进行增强的目的;最后,将低频图像与处理后的高频图像进行融合得到增强图像。实验结果证明,所提算法有效改善了磁瓦表面缺陷图像的光照不均匀、提高了图像的对比度、增强了图像的纹理细节信息,使得图像的缺陷部分和细节部分更加清晰可见。展开更多
To solve the problems of noise,detail loss and poor contrast in the successive mean quantization transform(SMQT),a new SMQT algorithm based on Otsu algorithm is proposed.In this algorithm,we integrate the optimal th...To solve the problems of noise,detail loss and poor contrast in the successive mean quantization transform(SMQT),a new SMQT algorithm based on Otsu algorithm is proposed.In this algorithm,we integrate the optimal threshold selected by the Otsu algorithm into the SMQT algorithm,then obtain the successive mean quantization of the binary tree.By this algorithm,an enhanced image is output with a higher quality.From both subjective visual effect and objective quality evaluation,the experimental results show that the improved algorithm reduces noise,improves contrast and makes the image details more clear.展开更多
文摘针对磁瓦表面缺陷图像整体光照不均匀、对比度较低、纹理细节信息不清晰等问题,提出了一种改进的同态滤波与连续均值量化变换(successive mean quantization transform,SMQT)相结合的磁瓦表面缺陷图像增强算法。首先,使用改进的巴特沃斯传递函数替换传统同态滤波中的传递函数,使其将图像分解成高频图像和低频图像;然后,使用SMQT算法非线性拉伸高频图像的灰度范围使其扩展到整个灰度级区域,从而达到对高频图像进行增强的目的;最后,将低频图像与处理后的高频图像进行融合得到增强图像。实验结果证明,所提算法有效改善了磁瓦表面缺陷图像的光照不均匀、提高了图像的对比度、增强了图像的纹理细节信息,使得图像的缺陷部分和细节部分更加清晰可见。
基金Supported by the National Natural Science Foundation of China(61503289)Hubei Province Science and Technology Support Program(2015BAA120,2015BCE068)
文摘To solve the problems of noise,detail loss and poor contrast in the successive mean quantization transform(SMQT),a new SMQT algorithm based on Otsu algorithm is proposed.In this algorithm,we integrate the optimal threshold selected by the Otsu algorithm into the SMQT algorithm,then obtain the successive mean quantization of the binary tree.By this algorithm,an enhanced image is output with a higher quality.From both subjective visual effect and objective quality evaluation,the experimental results show that the improved algorithm reduces noise,improves contrast and makes the image details more clear.