一生思破红尘路
剑藏庐轩隐迷踪
前处理
通过双边滤波去除高斯噪声。
使用opencv实现：
cv:: bilateralFilter(src_img, temp, 25, 25 * 2, 25 / 2);
自适应维纳滤波
通过自适应维纳滤波处理运动模糊。
维纳滤波实现代码如下（opencv）：
double WienerFilterImpl(const cv::Mat &src, cv::Mat &dst, double noiseVariance, const cv::Size &block) {
assert(("Invalid block dimensions", block.width % 2 == 1 && block.height % 2 == 1 && block.width > 1 &&
block.height > 1));
assert(("src and dst must be one channel grayscale images", src.channels() == 1, dst.channels() == 1));
int h = src.rows;
int w = src.cols;
dst = cv::Mat1b(h, w);
cv::Mat1d means, sqrMeans, variances;
cv::Mat1d avgVarianceMat;
cv::boxFilter(src, means, CV_64F, block, cv::Point(-1, -1), true, cv::BORDER_REPLICATE);
cv::sqrBoxFilter(src, sqrMeans, CV_64F, block, cv::Point(-1, -1), true, cv::BORDER_REPLICATE);
cv::Mat1d means2 = means.mul(means);
variances = sqrMeans - (means.mul(means));
if (noiseVariance < 0) {
cv::reduce(variances, avgVarianceMat, 1, CV_REDUCE_SUM, -1);
cv::reduce(avgVarianceMat, avgVarianceMat, 0, CV_REDUCE_SUM, -1);
noiseVariance = avgVarianceMat(0, 0) / (h * w);
}
for (int r = 0; r < h; ++r) {
uchar const *const srcRow = src.ptr<uchar>(r);
uchar *const dstRow = dst.ptr<uchar>(r);
double *const varRow = variances.ptr<double>(r);
double *const meanRow = means.ptr<double>(r);
for (int c = 0; c < w; ++c) {
dstRow[c] = cv::saturate_cast<uchar>(
meanRow[c] +
std::max(0., varRow[c] - noiseVariance) / std::max(varRow[c], noiseVariance) *
(srcRow[c] - meanRow[c])
);
}
}
return noiseVariance;
}
void WienerFilter(const cv::Mat &src, cv::Mat &dst, double noiseVariance = 10000, const cv::Size &block = cv::Size(3, 3)) {
WienerFilterImpl(src, dst, noiseVariance, block);
return ;
}
效果展示
处理前：
处理后：
