摘要
工业中用于无损检测的X光图像存在信噪比低、清晰度差、对比度低等缺点。为改善X光图像质量,提出一种改进的多尺度Retinex滤波(MSR)算法。将原始图像进行小波分解,在低频子带上进行MSR滤波,在高频子带上进行阈值滤波,实现图像重构。实验结果表明,改进MSR算法能同时达到图像增强和去噪的效果。
The X-ray films for nondestructive flaw detection in industry have the character of low Signal to Noise Ratio(SNR), bad definition and contrast. In order to improve the quality of the X-ray images, an improved Multi-scale Retinex(MSR) algorithm is proposed in this paper. Wavelet transforms the original image and MSR is filtered in the low frequency band. It offers a new threshold function and threshold filtering in the high frequency band to reconstruct the image. Experimental result shows that this MSR algorithm can efficiently enhance the image enhancement while reduce its noise.
出处
《计算机工程》
CAS
CSCD
2012年第8期186-188,共3页
Computer Engineering
关键词
图像增强
多尺度Retinex滤波
小波分解
阈值滤波
信噪比
阈值分割
image enhancement
Multi-scale Retinex(MSR) filtering
wavelet decomposition
threshold filtering
Signal to Noise Ratio(SNR)
threshold segmentation