# -*- coding: utf-8 -*-
"""
Created on Wed Aug 22 16:53:16 2018

@author: lenovo
"""

'''
SIFT算法
'''
import cv2

'''1、加载图片'''
img1 = cv2.imread('./image/match1.jpg')
img1 = cv2.resize(img1,dsize=(600,400))
gray1 = cv2.cvtColor(img1,cv2.COLOR_BGR2GRAY)
img2 = cv2.imread('./image/match2.jpg')
img2 = cv2.resize(img2,dsize=(600,400))
gray2 = cv2.cvtColor(img2,cv2.COLOR_BGR2GRAY)
image1 = gray1.copy()
image2 = gray2.copy()

'''2、提取特征点'''
#创建一个SIFT对象
sift = cv2.xfeatures2d.SIFT_create(400)
#SIFT对象会使用DoG检测关键点，并且对每个关键点周围的区域计算特征向量。该函数返回关键点的信息和描述符
keypoints1,descriptor1 = sift.detectAndCompute(image1,None)
keypoints2,descriptor2 = sift.detectAndCompute(image2,None)
print('descriptor1:',descriptor1.shape,'descriptor2',descriptor2.shape)
#在图像上绘制关键点
image1 = cv2.drawKeypoints(image=image1,keypoints = keypoints1,outImage=image1,color=(255,0,255),flags=cv2.DRAW_MATCHES_FLAGS_DRAW_RICH_KEYPOINTS)
image2 = cv2.drawKeypoints(image=image2,keypoints = keypoints2,outImage=image2,color=(255,0,255),flags=cv2.DRAW_MATCHES_FLAGS_DRAW_RICH_KEYPOINTS)
#显示图像
cv2.imshow('sift_keypoints1',image1)
cv2.imshow('sift_keypoints2',image2)
cv2.waitKey(20)


'''3、特征点匹配'''
matcher = cv2.FlannBasedMatcher()
matchePoints = matcher.match(descriptor1,descriptor2)
print(type(matchePoints),len(matchePoints),matchePoints[0])

#提取强匹配特征点
minMatch = 1
maxMatch = 0
for i in range(len(matchePoints)):
    if  minMatch > matchePoints[i].distance:
        minMatch = matchePoints[i].distance
    if  maxMatch < matchePoints[i].distance:
        maxMatch = matchePoints[i].distance
print('最佳匹配值是:',minMatch)        
print('最差匹配值是:',maxMatch)      

#获取排雷在前边的几个最优匹配结果
goodMatchePoints = []
for i in range(len(matchePoints)):
    if matchePoints[i].distance < minMatch + (maxMatch-minMatch)/3:
        goodMatchePoints.append(matchePoints[i])

#绘制最优匹配点
outImg = None
outImg = cv2.drawMatches(img1,keypoints1,img2,keypoints2,goodMatchePoints,outImg,matchColor=(0,255,0),flags=cv2.DRAW_MATCHES_FLAGS_DEFAULT)
cv2.imshow('matche',outImg)
cv2.waitKey(0)
cv2.destroyAllWindows()