import cv2 as cv import numpy as np # 加载与现实图像 src = cv.imread("D:/images/lena.jpg") cv.namedWindow("input", cv.WINDOW_AUTOSIZE) cv.imshow("input", src) # 转换为灰度 gray = cv.cvtColor(src, cv.COLOR_BGR2GRAY) cv.imshow("gray", gray) print(src.shape) print(gray.shape) cv.imwrite("D:/gray.png", gray) # 创建空白图像 black = np.zeros_like(src) cv.imshow("black", black) cv.imwrite("D:/black.png", black) # 调节亮度 black[:,:,:] = 50 lighter = cv.add(src, black) darker = cv.subtract(src, black) cv.imshow("lightness", lighter) cv.imshow("darkness", darker) cv.imwrite("D:/lightness.png", lighter) cv.imwrite("D:/darkness.png", darker) # 调节对比度 dst = cv.addWeighted(src, 1.2, black, 0.0, 0) cv.imshow("contrast", dst) cv.imwrite("D:/contrast.png", dst) # scale h, w, c = src.shape dst = cv.resize(src, (h//2, w//2)) cv.imshow("resize-image", dst) # 左右翻转 dst = cv.flip(src, 1) cv.imshow("flip", dst) # 上下翻转 dst = cv.flip(src, 0) cv.imshow("flip0", dst) cv.imwrite("D:/flip0.png", dst) # rotate M = cv.getRotationMatrix2D((w//2, h//2),45, 1) dst = cv.warpAffine(src, M, (w, h)) cv.imshow("rotate", dst) cv.imwrite("D:/rotate.png", dst) # 色彩 # HSV hsv = cv.cvtColor(src, cv.COLOR_BGR2HSV) cv.imshow("hsv", hsv) # 色彩表 - 支持14种色彩变换 dst = cv.applyColorMap(src, cv.COLORMAP_AUTUMN) cv.imshow("color table", dst) cv.imwrite("D:/color_table.png", dst) # blur blur = cv.blur(src, (15, 15)) cv.imshow("blur", blur) cv.imwrite("D:/blur.png", blur) # gaussian blur gblur = cv.GaussianBlur(src, (0, 0), 15) cv.imshow("gaussian blur", gblur) cv.imwrite("D:/gaussian.png", gblur) # custom filter - blur k = np.ones(shape=[5, 5], dtype=np.float32) / 25 dst = cv.filter2D(src, -1, k) cv.imshow("custom blur", dst) cv.imwrite("D:/custom_blur.png", dst) # EPF dst = cv.bilateralFilter(src, 0, 100, 10) cv.imshow("bi-filter", dst) cv.imwrite("D:/bi_blur.png", dst) # gradient dx = cv.Sobel(src, cv.CV_32F, 1, 0) dy = cv.Sobel(src, cv.CV_32F, 0, 1) dx = cv.convertScaleAbs(dx) dy = cv.convertScaleAbs(dy) cv.imshow("grad-x", dx) cv.imshow("grad-y", dy) cv.imwrite("D:/grad.png", dx) # edge detect edge = cv.Canny(src, 100, 300) cv.imshow("edge", edge) cv.imwrite("D:/edge.png", edge) # 直方图均衡化 eh = cv.equalizeHist(gray) cv.imshow("eh", eh) cv.imwrite("D:/eh.png", eh) # 角点检测 corners = cv.goodFeaturesToTrack(gray, 100, 0.05, 10) # print(len(corners)) for pt in corners: # print(pt) b = np.random.random_integers(0, 256) g = np.random.random_integers(0, 256) r = np.random.random_integers(0, 256) x = np.int32(pt[0][0]) y = np.int32(pt[0][1]) cv.circle(src, (x, y), 5, (int(b), int(g), int(r)), 2) cv.imshow("corners detection", src) cv.imwrite("D:/corners.png", src) # 二值化 src = cv.imread("D:/images/zsxq/zsxq_12.jpg") gray = cv.cvtColor(src, cv.COLOR_BGR2GRAY) cv.imshow("binary input", gray) # 固定阈值 ret, binary = cv.threshold(gray, 127, 255, cv.THRESH_BINARY) cv.imshow("binary", binary) cv.imwrite("D:/binary.png", binary) # 全局阈值 ret, binary = cv.threshold(gray, 0, 255, cv.THRESH_BINARY | cv.THRESH_OTSU) cv.imshow("otsu", binary) # 自适应阈值 binary = cv.adaptiveThreshold(gray, 255, cv.ADAPTIVE_THRESH_GAUSSIAN_C, cv.THRESH_BINARY_INV, 25, 10) cv.imshow("ada", binary) cv.imwrite("D:/ada.png", binary) # 轮廓分析 contours, hireachy = cv.findContours(binary, cv.RETR_EXTERNAL, cv.CHAIN_APPROX_SIMPLE) result = np.zeros_like(src) for cnt in range(len(contours)): cv.drawContours(result, contours, cnt, (0, 0, 255), 2, 8) cv.imshow("contour", result) cv.imwrite("D:/contour.png", result) # 膨胀与腐蚀操作 se = cv.getStructuringElement(cv.MORPH_RECT, (5, 5), (-1, -1)) d = cv.dilate(binary, se) e = cv.erode(binary, se) cv.imshow("dilate", d) cv.imshow("erode", e) # 开闭操作 op = cv.morphologyEx(binary, cv.MORPH_OPEN, se) cl = cv.morphologyEx(binary, cv.MORPH_CLOSE, se) cv.imshow("open", op) cv.imshow("close", cl) cv.waitKey(0) cv.destroyAllWindows()