python毕业设计：支持多边形选择区域的图像相似度比较软件的设计与实现+源代码+文档说明

# -*- coding: utf-8 -*- import time from tkinter import * from tkinter.filedialog import LoadFileDialog from tkinter.ttk import Separator import gray import color import texture import blob import cv2 import numpy as np from PIL import Image, ImageTk # -----------------------鼠标操作相关------------------------------------------ lsPointsChoose = [] tpPointsChoose = [] pointsCount = 0 count = 0 pointsMax = 6 img = 0 ROI = 0 def on_mouse(event, x, y, flags, param): global img, point1, point2, count, pointsMax, window_name global lsPointsChoose, tpPointsChoose # 存入选择的点 global pointsCount # 对鼠标按下的点计数 global img2, ROI_bymouse_flag img2 = img.copy() # 此行代码保证每次都重新再原图画 避免画多了 # ----------------------------------------------------------- # count=count+1 # print("callback_count",count) # -------------------------------------------------------------- if event == cv2.EVENT_LBUTTONDOWN: # 左键点击 pointsCount = pointsCount + 1 print('第' + str(pointsCount) + '次点击') # 感觉这里没有用？2018年8月25日20:06:42 # 为了保存绘制的区域，画的点稍晚清零 # if (pointsCount == pointsMax + 1): # pointsCount = 0 # tpPointsChoose = [] point1 = (x, y) # 画出点击的点 cv2.circle(img2, point1, 10, (0, 255, 0), 2) print(point1) # 将选取的点保存到list列表里 lsPointsChoose.append([x, y]) # 用于转化为darry 提取多边形ROI tpPointsChoose.append((x, y)) # 用于画点 # ---------------------------------------------------------------------- print('点的个数：', len(tpPointsChoose)) # 将所有的点都标红处理 for i in range(len(tpPointsChoose)): cv2.circle(img2, tpPointsChoose[i], 2, (255, 0, 0), 2) # 将鼠标选的点用直线连起来 for i in range(len(tpPointsChoose) - 1): print('i:', i, end=' --> ') cv2.line(img2, tpPointsChoose[i], tpPointsChoose[i + 1], (0, 0, 255), 2) print(tpPointsChoose[i], '————', tpPointsChoose[i + 1]) # ---------------------------------------------------------------------- ''' # ----------点击到pointMax时可以提取去绘图---------------- if (pointsCount == pointsMax): # -----------绘制感兴趣区域----------- ROI_byMouse() ROI_bymouse_flag = 1 i = 0 pointsCount = 0 tpPointsChoose = [] lsPointsChoose = [] ''' # ----------点击到tpPointsChoose[0]附近时时可以提取去绘图---------------- if (pointsCount > 2) and (x <= tpPointsChoose[0][0] + 3) and (x >= tpPointsChoose[0][0] - 3) and ( y <= tpPointsChoose[0][1] + 3) and (y >= tpPointsChoose[0][1] - 3): # -----------绘制感兴趣区域----------- ROI_byMouse() ROI_bymouse_flag = 1 i = 0 pointsCount = 0 tpPointsChoose = [] lsPointsChoose = [] cv2.destroyWindow(window_name) return cv2.imshow(window_name, img2) # -------------------------右键按下清除轨迹----------------------------- if event == cv2.EVENT_RBUTTONDOWN: # 右键点击 print("right-mouse") pointsCount = 0 tpPointsChoose = [] lsPointsChoose = [] print(len(tpPointsChoose)) for i in range(len(tpPointsChoose) - 1): print('i', i) cv2.line(img2, tpPointsChoose[i], tpPointsChoose[i + 1], (0, 0, 255), 2) cv2.imshow(window_name, img2) def ROI_byMouse(): global src, ROI, ROI_flag, mask2, lsPointsChoose, img_no mask = np.zeros(img.shape, np.uint8) print(lsPointsChoose) pts = np.array([lsPointsChoose], np.int32) # pts是多边形的顶点列表（顶点集） pts = pts.reshape((-1, 1, 2)) # 这里 reshape 的第一个参数为-1, 表明这一维的长度是根据后面的维度的计算出来的。 # OpenCV中需要先将多边形的顶点坐标变成顶点数×1×2维的矩阵，再来绘制 # --------------画多边形--------------------- mask = cv2.polylines(mask, [pts], True, (255, 255, 255)) ##-------------填充多边形--------------------- mask2 = cv2.fillPoly(mask, [pts], (255, 255, 255)) ''' cv2.imwrite('mask.jpg', mask2) cv2.imshow('mask', mask2) ''' ROI = cv2.bitwise_and(mask2, img) ''' cv2.imwrite('ROI.jpg', ROI) cv2.imshow('ROI', ROI) ''' # 截取规则图像 rect = cv2.minAreaRect(pts) box = np.int0(cv2.boxPoints(rect)) IMG = ROI[min(box[:, 1]):max(box[:, 1]), min(box[:, 0]):max(box[:, 0])] cv2.imwrite("IMG" + str(img_no) + ".jpg", IMG) # cv2.imshow('IMG', IMG) filename = "" window_name = "" def myopen(): global filename, img, ROI, window_name fd = LoadFileDialog(root) # 创建打开文件对话框 filename = fd.go() # 显示打开文件对话框，并获取选择的文件名称 print(filename) img = cv2.imread(filename) ROI = img.copy() window_name = "Picture" + str(img_no) cv2.namedWindow(window_name) cv2.setMouseCallback(window_name, on_mouse) cv2.imshow(window_name, img) cv2.waitKey(0) # cv2.destroyAllWindows() img_no = 0 # 当前所选的图片的编号，取值1或2，默认0 is_choose_1 = False is_choose_2 = False def myopen1(): global img_no, is_choose_1, lbl_img1, img1_new, cmp_msg img_no = 1 is_choose_1 = True myopen() img_t = Image.open("IMG1.jpg") img1_new = ImageTk.PhotoImage(img_t) lbl_img1.configure(image=img1_new) lbl_img1.image = img1_new cmp_msg.set("成功打开图片1") def myopen2(): global img_no, is_choose_2, lbl_img2, img2_new, cmp_msg img_no = 2 is_choose_2 = True myopen() img_t = Image.open("IMG2.jpg") img2_new = ImageTk.PhotoImage(img_t) lbl_img2.configure(image=img2_new) lbl_img2.image = img2_new cmp_msg.set("成功打开图片2") def result(): global cmp_msg if is_choose_1 == 0 and is_choose_2 == 0: cmp_msg.set("请先完成图片选择") return if is_choose_1 == 0: cmp_msg.set("请继续选择图片1") return if is_choose_2 == 0: cmp_msg.set("请继续选择图片2") return imgfile1 = cv2.imread("IMG1.jpg") imgfile2 = cv2.imread("IMG2.jpg") # 灰度相似度 sim_gray = gray.GrayComparison(imgfile1, imgfile2) # 色度相似度 sim_color = color.ColorComparison(imgfile1, imgfile2) # 纹理相似度 sim_texture = texture.TexrtureComparison(imgfile1, imgfile2) # 斑块相似度 sim_blob = blob.BlobComparison(imgfile1, imgfile2) cmp = "灰度相似度: " + str(round(sim_gray, 2)) + "\n色度相似度: " + str(round(sim_color, 2)) + "\n纹理相似度: " + str( round(sim_texture, 2)) + "\n斑块相似度: " + str(round(sim_blob, 2)) + "\n" cmp_msg.set(cmp) # GUI设计 root = Tk() # 建立根窗口，命名为root root.title("图像相似度比较软件——18074305沈海健") # 根窗口root设置标题 root.resizable(False, False) # 根窗口root不可改变大小 frm_name = Frame(root) frm_name.pack(side=TOP, anchor=CENTER) Label(frm_name, text="提示：所选区域越接近矩形相似度越准确").pack(side="left") Separator(root, orient=HORIZONTAL).pack(fill=X, padx=5) frm_choose = Frame(root) frm_choose.pack(side=TOP) Button(frm_choose, text="打开图片1", command=myopen1).pack(side=LEFT) Button(frm_choose, text="打开图片