OpenCV开发笔记（七十八）：在ubuntu上搭建opencv+python开发环境以及匹配识别Demo

news2024/9/19 9:27:33

若该文为原创文章，转载请注明原文出处
本文章博客地址：https://hpzwl.blog.csdn.net/article/details/140435870

长沙红胖子Qt（长沙创微智科）博文大全：开发技术集合（包含Qt实用技术、树莓派、三维、OpenCV、OpenGL、ffmpeg、OSG、单片机、软硬结合等等）持续更新中…

OpenCV开发专栏（点击传送门）

上一篇：《OpenCV开发笔记（七十七）：相机标定（二）：通过棋盘标定计算相机内参矩阵矫正畸变摄像头图像》
下一篇：持续补充中…

前言

Python上的OpenCv开发，在linux上的基本环境搭建流程。

安装python

以python2.7为开发版本。

sudo apt-get install python2.7
sudo apt-get install python2.7-dev

安装OpenCV

多种方式，先选择最简单的方式。

sudo apt-get install python-opencv

打开摄像头

测试Demo

import cv2
import numpy
cap = cv2.VideoCapture(0)
while 1:
  ret, frame = cap.read()
  cv2.imshow("capture", frame)
  if cv2.waitKey(100) & 0xff == ord('q'):
    break
cap.release()
cv2.destroyAllWindows()

在这里插入图片描述

测试结果

在这里插入图片描述

模板匹配

测试Demo

import cv2
import numpy

# read template image
template = cv2.imread("src.png")
#cv2.imshow("template", template);

# read target image
target = cv2.imread("dst.png")
#cv2.imshow("target", target)

# get tempalte's width and height
tHeight, tWidth = template.shape[:2]
print tHeight, tWidth

# matches
result = cv2.matchTemplate(target, template, cv2.TM_SQDIFF_NORMED)

# normalize
cv2.normalize(result, result, 0, 1, cv2.NORM_MINMAX, -1)

minVal, maxVal, minLoc, maxLoc = cv2.minMaxLoc(result)

strminVal = str(minVal)
print strminVal

cv2.rectangle(target, minLoc, (minLoc[0] + tWidth, minLoc[1] + tHeight), (0,0,255), 2)

cv2.imshow("result", target)

cv2.waitKey()

cv2.destroyAllWindows()

测试结果

在这里插入图片描述

Flann特征点匹配

版本回退

在opencv3.4.x大版本后，4.x系列的sift被申请了专利，无法使用了，flann需要使用到

sift = cv2.xfeatures2d.SIFT_create()

所以需要回退版本。

sudo apt-get remove python-opencv
sudo pip install opencv-python==3.4.2.16

安装模块库matplotlib

python -m pip install matplotlib
sudo apt-get install python-tk
pip install opencv-contrib-python==3.4.2.16

测试Demo

# FLANN based Matcher
import numpy as np
import cv2
from matplotlib import pyplot as plt
 
#min match count is 10
MIN_MATCH_COUNT = 10 
# queryImage
template = cv2.imread('src.png',0) 
# trainImage
target = cv2.imread('dst.png',0) 
# initiate SIFT detector
sift = cv2.xfeatures2d.SIFT_create()
# find the keypoints and descriptors with SIFT
kp1, des1 = sift.detectAndCompute(template,None)
kp2, des2 = sift.detectAndCompute(target,None)
# create FLANN match
FLANN_INDEX_KDTREE = 0
index_params = dict(algorithm = FLANN_INDEX_KDTREE, trees = 5)
search_params = dict(checks = 50)
flann = cv2.FlannBasedMatcher(index_params, search_params)
matches = flann.knnMatch(des1,des2,k=2)
# store all the good matches as per Lowe's ratio test.
good = []
# lose < 0.7
for m,n in matches:
    if m.distance < 0.7*n.distance:
        good.append(m)
if len(good)>MIN_MATCH_COUNT:
    # get key
    src_pts = np.float32([ kp1[m.queryIdx].pt for m in good ]).reshape(-1,1,2)
    dst_pts = np.float32([ kp2[m.trainIdx].pt for m in good ]).reshape(-1,1,2)
    # cal mat and mask
    M, mask = cv2.findHomography(src_pts, dst_pts, cv2.RANSAC, 5.0)
    matchesMask = mask.ravel().tolist()
    h,w = template.shape
    # convert 4 corner
    pts = np.float32([ [0,0],[0,h-1],[w-1,h-1],[w-1,0] ]).reshape(-1,1,2)
    dst = cv2.perspectiveTransform(pts,M)
    cv2.polylines(target,[np.int32(dst)],True,0,2, cv2.LINE_AA)
else:
    print( "Not enough matches are found - %d/%d" % (len(good),MIN_MATCH_COUNT))
    matchesMask = None
draw_params = dict(matchColor=(0,255,0), 
                   singlePointColor=None,
                   matchesMask=matchesMask, 
                   flags=2)

result = cv2.drawMatches(template, kp1, target, kp2, good, None, **draw_params)
cv2.imshow("dst", result)
cv2.imshow("dst2", target)
cv2.waitKey()