基于python+opencv的连铸漏钢感兴趣区域快速识别与标记源码+超详细注释(阈值分割、形态学操作、特征提取等).zip

# -*- coding: utf-8 -*- #!/usr/bin/env python # @File : batch_extraction.py # @Software: PyCharm import os import cv2 import numpy as np def process_picture(event, x, y, flags, param): if event == cv2.EVENT_LBUTTONDOWN: print("单击了鼠标左键") elif event == cv2.EVENT_RBUTTONDOWN : print("单击了鼠标右键") elif flags == cv2.EVENT_FLAG_LBUTTON: print("按住左键拖动了鼠标") elif event == cv2.EVENT_MBUTTONDOWN : print("单击了中间键") def picture_contours(o_img): """ 提取V字特征图像特征轮廓 :param o_img: 要处理的图片 :return: 全部的轮廓特征 """ # print("原图类型%s" % str(o_img.shape)) gray = cv2.cvtColor(o_img, cv2.COLOR_BGRA2GRAY) ret, binary = cv2.threshold(gray, 127, 255, cv2.THRESH_BINARY_INV) # cv2.imshow("binary-img", binary) cts, hierarchy = cv2.findContours(binary, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE) return cts # def intercept_contour(p_contours, index, c_contours): # n = len(contours) # # 原图像大小相同的画布 # contours_img = [] # max_area = 0 # max_i = 0 # for i in range(n): # temp = np.zeros(o_img.shape, np.uint8) # contours_img.append(temp) # contours_img[i] = cv2.drawContours(contours_img[i], contours, i, (255, 255, 255), -1) # if cv2.contourArea(contours[i]) > max_area: # max_area = cv2.contourArea(contours[i]) # max_i = i # # 获取最大轮廓的矩形边界值 # x, y, w, h = cv2.boundingRect(contours[max_i]) # # 画矩形 # cv2.rectangle(o_img, (x, y), (x + w, y + h), (0, 255, 0), 2) # # 取最小矩形轮廓 # # rect = cv2.minAreaRect(contours[max_i]) # # # 矩形四个角点取整 # # box = np.int0(cv2.boxPoints(rect)) # # cv2.drawContours(o, [box], 0, (0, 255, 0), 2) # # cv2.imshow("feature-img", contours_img[max_i]) # cv2.imshow("finally-img", o_img) # # 截取图片并保存（截取面积最大的） # if cv2.imwrite(os.path.join(s_path, p_name), o_img[y:y + h, x:x + w]): # print("保存图像V特征%s", p_name) # print("最大轮廓面积%s" % max_area) if __name__ == "__main__": # 要处理的图片路径 process_path = "G:\\MyProjects\\python\\Breakout-Identification\\morphologyEX" # 要保存的图片路径 save_path = "G:\\MyProjects\\python\\Breakout-Identification\\v_feature" os.chdir(process_path) print("当前路径%s" % os.getcwd()) print("当前路径下的文件%s" % os.listdir(os.getcwd())) # 图片路径列表 picture_path_list = os.listdir(os.getcwd()) # 排除不是png图片的文件或文件夹 picture_path_list = [picture_path_list[i] for i in range(0, len(picture_path_list)) if picture_path_list[i].find('png') != -1] # 图片个数 picture_list_number = len(picture_path_list) # 图片路径列表的索引 picture_path_index = 0 # 轮廓索引 c_index = 0 # 当前轮廓索引 n_index = 0 # 原图像大小相同的画布 contours_img = [] # 实时显示的轮廓索引列表 n_s_index_list = [] # 要保存的轮廓图像矩阵 save_temp = [] while (1): # 图片索引超出时跳出循环 if picture_path_index > picture_list_number-1: print("图片已经全部处理完") break # 读取图片 img = cv2.imread(os.path.join(process_path, picture_path_list[picture_path_index])) # 图片文件名 img_name = picture_path_list[picture_path_index] # 图片的全部轮廓 contours = picture_contours(img) # 轮廓个数 c_num = len(contours) # 等待键盘输入 k_value = cv2.waitKey(1) & 0xFF # 显示原图 cv2.imshow(img_name, img) # 实时显示要保存的特征图 if n_s_index_list: # 将保存的轮廓放一块显示 temp = np.zeros(img.shape, np.uint8) for i in n_s_index_list: temp = cv2.drawContours(temp, contours, i, (255, 255, 255), -1) save_temp = temp cv2.imshow("save-contour", temp) else: cv2.imshow("save-contour", np.zeros(img.shape, np.uint8)) # 实时显示单个轮廓图片 if contours_img: cv2.imshow("single-contour", contours_img[n_index]) else: cv2.imshow("single-contour", np.zeros(img.shape, np.uint8)) # 读取后一张后一张 if k_value == ord('d'): picture_path_index += 1 c_index = 0 contours_img.clear() n_s_index_list.clear() cv2.destroyAllWindows() print("后一张") # 保存图片V特征 elif k_value == ord('s'): ret, binary = cv2.threshold(save_temp, 127, 255, cv2.THRESH_BINARY_INV) loc = cv2.bitwise_or(img, binary) if cv2.imwrite(os.path.join(save_path, img_name), loc): print("保存V字特征轮廓+1") # 查看一个轮廓的特征 elif k_value == ord('n'): if c_index == c_num: print("已显示全部轮廓") c_index = 0 contours_img.clear() n_s_index_list.clear() cv2.destroyAllWindows() continue temp = np.zeros(img.shape, np.uint8) contours_img.append(temp) cv2.drawContours(contours_img[c_index], contours, c_index, (255, 255, 255), -1) n_index = c_index c_index += 1 # 记录要保存的轮廓索引 elif k_value == ord('m'): # 添加保存轮廓的在contours的索引 n_s_index_list.append(n_index) print("记录当前轮廓") # 按ESC退出 elif k_value == 27: break cv2.destroyAllWindows()