import math import cv2 import numpy as np import struct import matplotlib.pyplot as plt from PyQt5.QtWidgets import QMessageBox class EdgeMenu: def __init__(self, pic_path): self.pic_path = pic_path self.pic = cv2.imread(self.pic_path, 0).astype(np.uint8) #灰度模式读取 self.cpic = cv2.imread(self.pic_path, 1).astype(np.uint8) def ShowPic(self): cv2.imshow('当前图片', self.pic) cv2.waitKey(0) # 0一直显示，直到有键盘输入。也可以是其他数字. def PicStrength(self): self.pic = self.pic.astype('float') row, column = self.pic.shape gradient = np.zeros((row, column)) for x in range(row - 1): for y in range(column - 1): gx = abs(self.pic[x + 1, y] - self.pic[x, y]) gy = abs(self.pic[x, y + 1] - self.pic[x, y]) gradient[x, y] = gx + gy # 3. 对图像进行增强，增强后的图像变量名为sharp sharp = self.pic + gradient sharp = np.where(sharp > 255, 255, sharp) sharp = np.where(sharp < 0, 0, sharp) gradient = gradient.astype('uint8') sharp = sharp.astype('uint8') cv2.imwrite('img/EdgeImgs/EdgePic.jpg', sharp) cv2.waitKey(0) def EdgeRoberts(self): kernelx = np.array([[-1, 0], [0, 1]], dtype=int) kernely = np.array([[0, -1], [1, 0]], dtype=int) x = cv2.filter2D(self.pic, cv2.CV_16S, kernelx) y = cv2.filter2D(self.pic, cv2.CV_16S, kernely) absX = cv2.convertScaleAbs(x) absY = cv2.convertScaleAbs(y) Roberts = cv2.addWeighted(absX, 0.5, absY, 0.5, 0) cv2.imwrite('img/EdgeImgs/EdgePic.jpg', Roberts) cv2.waitKey(0) def EdgeSobel(self): x = cv2.Sobel(self.pic, cv2.CV_16S, 1, 0) y = cv2.Sobel(self.pic, cv2.CV_16S, 0, 1) absX = cv2.convertScaleAbs(x) absY = cv2.convertScaleAbs(y) Sobel = cv2.addWeighted(absX, 0.5, absY, 0.5, 0) cv2.imwrite('img/EdgeImgs/EdgePic.jpg', Sobel) cv2.waitKey(0) def EdgeLaplacian(self, k_size): grayImage = cv2.GaussianBlur(self.pic, (5, 5), 0, 0) # 3. 拉普拉斯算法 if k_size == '1': dst = cv2.Laplacian(grayImage, cv2.CV_16S, ksize=1) elif k_size == '3': dst = cv2.Laplacian(grayImage, cv2.CV_16S, ksize=3) elif k_size == '5': dst = cv2.Laplacian(grayImage, cv2.CV_16S, ksize=5) elif k_size == '7': dst = cv2.Laplacian(grayImage, cv2.CV_16S, ksize=7) else: return # 4. 数据格式转换 Laplacian = cv2.convertScaleAbs(dst) cv2.imwrite('img/EdgeImgs/EdgePic.jpg', Laplacian) cv2.waitKey(0) def EdgeLoG(self): img = cv2.imread(self.pic_path) grayImage = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) # 2. 边缘扩充处理图像 image = cv2.copyMakeBorder(grayImage, 2, 2, 2, 2, borderType=cv2.BORDER_REPLICATE) image = cv2.GaussianBlur(image, (3, 3), 0, 0) # 3. 使用Numpy定义LoG算子 m1 = np.array( [[0, 0, -1, 0, 0], [0, -1, -2, -1, 0], [-1, -2, 16, -2, -1], [0, -1, -2, -1, 0], [0, 0, -1, 0, 0]]) rows = image.shape[0] cols = image.shape[1] image1 = np.zeros(image.shape) # 4. 卷积运算 for k in range(0, 2): for i in range(2, rows - 2): for j in range(2, cols - 2): image1[i, j] = np.sum((m1 * image[i - 2:i + 3, j - 2:j + 3, k])) image1 = cv2.convertScaleAbs(image1) cv2.imwrite('img/EdgeImgs/EdgePic.jpg', image1) cv2.waitKey(0) def EdgeCanny(self): self.picColor = cv2.imread(self.pic_path, 1).astype(np.uint8) #彩色模式读取 # 2. 灰度转换 blur = cv2.cvtColor(self.picColor, cv2.COLOR_BGR2GRAY) # 3. 求x，y方向的Sobel算子 gradx = cv2.Sobel(blur, cv2.CV_16SC1, 1, 0) grady = cv2.Sobel(blur, cv2.CV_16SC1, 0, 1) # 4. 使用Canny函数处理图像，x,y分别是3求出来的梯度，低阈值50，高阈值150 edge_output = cv2.Canny(gradx, grady, 50, 150) cv2.imwrite('img/EdgeImgs/EdgePic.jpg', edge_output) cv2.waitKey(0) def HoughDetect(self, threshold, color): img = cv2.GaussianBlur(self.cpic, (3, 3), 0) edges = cv2.Canny(img, 50, 150, apertureSize=3) lines = cv2.HoughLines(edges, 1, np.pi / 2, threshold) result = img.copy() if lines is None: cv2.imwrite('img/EdgeImgs/EdgePic.jpg', result) return for i_line in lines: for line in i_line: rho = line[0] theta = line[1] if (theta < (np.pi / 4.)) or (theta > (3. * np.pi / 4.0)): # 垂直直线 pt1 = (int(rho / np.cos(theta)), 0) pt2 = (int((rho - result.shape[0] * np.sin(theta)) / np.cos(theta)), result.shape[0]) if color == 'red': cv2.line(result, pt1, pt2, (0, 0, 255)) if color == 'white': cv2.line(result, pt1, pt2, (255, 255, 255)) if color == 'black': cv2.line(result, pt1, pt2, (0, 0, 0)) if color == 'green': cv2.line(result, pt1, pt2, (0, 255, 0)) if color == 'blue': cv2.line(result, pt1, pt2, (0, 255, 0)) else: pt1 = (0, int(rho / np.sin(theta))) pt2 = (result.shape[1], int((rho - result.shape[1] * np.cos(theta)) / np.sin(theta))) if color == 'red': cv2.line(result, pt1, pt2, (0, 0, 255), 1) if color == 'white': cv2.line(result, pt1, pt2, (255, 255, 255), 1) if color == 'black': cv2.line(result, pt1, pt2, (0, 0, 0), 1) if color == 'green': cv2.line(result, pt1, pt2, (0, 255, 0), 1) if color == 'blue': cv2.line(result, pt1, pt2, (0, 255, 0), 1) cv2.imwrite('img/EdgeImgs/EdgePic.jpg', result) cv2.waitKey(0) def HoughDetectP(self, threshold, MinLineLength, MaxLineGap, color): img = cv2.GaussianBlur(self.cpic, (3, 3), 0) edges = cv2.Canny(img, 50, 150, apertureSize=3) linesP = cv2.HoughLinesP(edges, 1, np.pi / 180, threshold, minLineLength = MinLineLength,maxLineGap = MaxLineGap) result_P = img.copy() if linesP is None: cv2.imwrite('img/EdgeImgs/EdgePic.jpg', result_P) return for i_P in linesP: for x1, y1, x2, y2 in i_P: if color == 'red': cv2.line(result_P, (x1, y1), (x2, y2), (0, 0, 255), 3) if color == 'white': cv2.line(result_P, (x1, y1), (x2, y2), (255, 255, 255), 3) if color == 'black': cv2.line(result_P, (x1, y1), (x2, y2), (0, 0, 0), 3) if color == 'green': cv2.line(result_P, (x1, y1), (x2, y2), (0, 255, 0), 3) if color == 'blue': cv2.line(result_P, (x1, y1), (x2, y2), (255, 0, 0), 3) cv2.imwrite('img/EdgeImgs/EdgePic.jpg', result_P) cv2.waitKey(0)