Zi 字媒體

fanfuhan OpenCV 教學038 ~ opencv-038-拉普拉斯金字塔 資料來源: https://fanfuhan.github.io/ https://fanfuhan.github.io/2019/04/11/opencv-038/ GITHUB:https://github.com/jash-git/fanfuhan_ML_OpenCV 知識點: 對輸入圖像實現金字塔的reduce操作就會生成不同分辨率的圖像、對這些圖像進行金字塔expand操作，然後使用reduce減去expand之後的結果就會得到圖像拉普拉斯金字塔圖像。 舉例如下： 輸入圖像G(0) 金字塔reduce操作生成G(1), G(2), G(3) 拉普拉斯金字塔： L0 = G(0)-expand(G(1)) L1 = G (1)-expand(G(2)) L2 = G(2)–expand(G(3)) G(0)減去expand(G(1))得到的結果就是兩次高斯模糊輸出的不同，所以L0稱為DOG（高斯不同）、它約等於LOG所以又稱為拉普拉斯金字塔。所以要求的圖像的拉普拉斯金字塔，首先要進行金字塔的reduce操作，然後在通過expand操作，最後相減得到拉普拉斯金字塔圖像。 C++ #include #include using namespace std; using namespace cv; void pyramid_up(Mat &image, vector &pyramid_images, int level); void laplaian_demo(vector &pyramid_images, Mat &image); /* * 拉普拉斯金字塔 */ int main() { Mat src = imread("../images/test1.jpg"); if (src.empty()) { cout << "could not load image.." << endl; } imshow("input", src); vector p_images; pyramid_up(src, p_images, 2); laplaian_demo(p_images, src); waitKey(0); return 0; } void laplaian_demo(vector &pyramid_images, Mat &image) { for (int i = pyramid_images.size() - 1; i > -1; --i) { Mat dst; if (i - 1 < 0) { pyrUp(pyramid_images[i], dst, image.size()); subtract(image, dst, dst); dst = dst + Scalar(127, 127, 127);# 调亮度， 实际中不能这么用 imshow(format("laplaian_layer_%d", i), dst); } else { pyrUp(pyramid_images[i], dst, pyramid_images[i-1].size()); subtract(pyramid_images[i - 1], dst, dst); dst = dst + Scalar(127, 127, 127); imshow(format("laplaian_layer_%d", i), dst); } } } void pyramid_up(Mat &image, vector &pyramid_images, int level) { Mat temp = image.clone(); Mat dst; for (int i = 0; i < level; ++i) { pyrDown(temp, dst); //imshow(format("pyramid_up_%d", i), dst); temp = dst.clone(); pyramid_images.push_back(temp); } } Python import cv2 as cv import numpy as np def laplaian_demo(pyramid_images): level = len(pyramid_images) for i in range(level-1, -1, -1): if (i-1) < 0: h, w = src.shape[:2] expand = cv.pyrUp(pyramid_images[i], dstsize=(w, h)) lpls = cv.subtract(src, expand) + 127 cv.imshow("lpls_" + str(i), lpls) else: h, w = pyramid_images[i-1].shape[:2] expand = cv.pyrUp(pyramid_images[i], dstsize=(w, h)) lpls = cv.subtract(pyramid_images[i-1], expand) + 127 cv.imshow("lpls_"+str(i), lpls) def pyramid_up(image, level=3): temp = image.copy() # cv.imshow("input", image) pyramid_images = [] for i in range(level): dst = cv.pyrDown(temp) pyramid_images.append(dst) # cv.imshow("pyramid_up_" + str(i), dst) temp = dst.copy() return pyramid_images src = cv.imread("D:/images/master.jpg") cv.namedWindow("input", cv.WINDOW_AUTOSIZE) cv.imshow("input", src) # pyramid_up(src) laplaian_demo(pyramid_up(src)) cv.waitKey(0) cv.destroyAllWindows()