devbox
Cpp五条
这个屌丝很懒，什么也没留下！
关注作者
热门标签
热门文章
当前位置:   article > 正文

Python cv2.minMaxLoc方法代码示例

作者：Cpp五条 | 2024-02-13 00:42:50

赞

踩

cv2.minmaxloc

本文整理汇总了Python中cv2.minMaxLoc方法的典型用法代码示例。如果您正苦于以下问题：Python cv2.minMaxLoc方法的具体用法？Python cv2.minMaxLoc怎么用？Python cv2.minMaxLoc使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在模块cv2的用法示例。

在下文中一共展示了cv2.minMaxLoc方法的22个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于我们的系统推荐出更棒的Python代码示例。

示例1: match_img

​ 点赞 7 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def match_img(image, template, value):
  4.     """
  5.     :param image: 图片
  6.     :param template: 模板
  7.     :param value: 阈值
  8.     :return: 水印坐标
  9.     描述:用于获得这幅图片模板对应的位置坐标,用途:校准元素位置信息
  10.     """
  11.     res = cv2.matchTemplate(image, template, cv2.TM_CCOEFF_NORMED)
  12.     threshold = value
  13.     min_v, max_v, min_pt, max_pt = cv2.minMaxLoc(res)
  14.     if max_v < threshold:
  15.         return False
  16.     if not max_pt[0] in range(10, 40) or max_pt[1] > 20:
  17.         return False
  18.     return max_pt

开发者ID:Mingtzge，项目名称:2019-CCF-BDCI-OCR-MCZJ-OCR-IdentificationIDElement，代码行数:18，代码来源:split_img_generate_data.py


 

 

示例2: get_match_confidence

​ 点赞 7 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def get_match_confidence(img1, img2, mask=None):
  4.     if img1.shape != img2.shape:
  5.         return False
  6.     ## first try, using absdiff
  7.     # diff = cv2.absdiff(img1, img2)
  8.     # h, w, d = diff.shape
  9.     # total = h*w*d
  10.     # num = (diff<20).sum()
  11.     # print 'is_match', total, num
  12.     # return num > total*0.90
  13.     if mask is not None:
  14.         img1 = img1.copy()
  15.         img1[mask!=0] = 0
  16.         img2 = img2.copy()
  17.         img2[mask!=0] = 0
  18.     ## using match
  19.     match = cv2.matchTemplate(img1, img2, cv2.TM_CCOEFF_NORMED)
  20.     _, confidence, _, _ = cv2.minMaxLoc(match)
  21.     # print confidence
  22.     return confidence

开发者ID:NetEaseGame，项目名称:ATX，代码行数:22，代码来源:scene_detector.py


 

示例3: probability

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def probability(self, im: str) -> float:
  4.         """
  5.         Return the probability of the existence of given image.
  7.         :param im: the name of the image.
  8.         :return: the probability (confidence).
  9.         """
  10.         assert self.screen is not None
  11.         try:
  12.             template = self.images[im]
  13.         except KeyError:
  14.             logger.error('Unexpected image name {}'.format(im))
  15.             return 0.0
  16.  
  17.         res = cv.matchTemplate(self.screen, template, TM_METHOD)
  18.         _, max_val, _, max_loc = cv.minMaxLoc(res)
  19.         logger.debug('max_val = {}, max_loc = {}'.format(max_val, max_loc))
  20.         return max_val

开发者ID:will7101，项目名称:fgo-bot，代码行数:20，代码来源:tm.py


 

 

示例4: find

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def find(self, im: str, threshold: float = None) -> Tuple[int, int]:
  4.         """
  5.         Find the template image on screen and return its top-left coords.
  7.         Return None if the matching value is less than `threshold`.
  9.         :param im: the name of the image
  10.         :param threshold: the threshold of matching. If not given, will be set to the default threshold.
  11.         :return: the top-left coords of the result. Return (-1, -1) if not found.
  12.         """
  13.         threshold = threshold or self.threshold
  14.  
  15.         assert self.screen is not None
  16.         try:
  17.             template = self.images[im]
  18.         except KeyError:
  19.             logger.error('Unexpected image name {}'.format(im))
  20.             return -1, -1
  21.  
  22.         res = cv.matchTemplate(self.screen, template, TM_METHOD)
  23.         _, max_val, _, max_loc = cv.minMaxLoc(res)
  24.         logger.debug('max_val = {}, max_loc = {}'.format(max_val, max_loc))
  25.         return max_loc if max_val >= threshold else (-1, -1)

开发者ID:will7101，项目名称:fgo-bot，代码行数:25，代码来源:tm.py


 

示例5: cal_rgb_confidence

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def cal_rgb_confidence(img_src_rgb, img_sch_rgb):
  4.     """同大小彩图计算相似度."""
  5.     # BGR三通道心理学权重:
  6.     weight = (0.114, 0.587, 0.299)
  7.     src_bgr, sch_bgr = cv2.split(img_src_rgb), cv2.split(img_sch_rgb)
  8.  
  9.     # 计算BGR三通道的confidence，存入bgr_confidence:
  10.     bgr_confidence = [0, 0, 0]
  11.     for i in range(3):
  12.         res_temp = cv2.matchTemplate(src_bgr[i], sch_bgr[i], cv2.TM_CCOEFF_NORMED)
  13.         min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res_temp)
  14.         bgr_confidence[i] = max_val
  15.  
  16.     # 加权可信度
  17.     weighted_confidence = bgr_confidence[0] * weight[0] + bgr_confidence[1] * weight[1] + bgr_confidence[2] * weight[2]
  18.  
  19.     return weighted_confidence

开发者ID:AirtestProject，项目名称:Airtest，代码行数:19，代码来源:cal_confidence.py


 

 

示例6: find_template

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def find_template(im_source, im_search, threshold=0.8, rgb=False):
  4.     """函数功能：找到最优结果."""
  5.     # 第一步：校验图像输入
  6.     check_source_larger_than_search(im_source, im_search)
  7.     # 第二步：计算模板匹配的结果矩阵res
  8.     res = _get_template_result_matrix(im_source, im_search)
  9.     # 第三步：依次获取匹配结果
  10.     min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)
  11.     h, w = im_search.shape[:2]
  12.     # 求取可信度:
  13.     confidence = _get_confidence_from_matrix(im_source, im_search, max_loc, max_val, w, h, rgb)
  14.     # 求取识别位置: 目标中心 + 目标区域:
  15.     middle_point, rectangle = _get_target_rectangle(max_loc, w, h)
  16.     best_match = generate_result(middle_point, rectangle, confidence)
  17.     LOGGING.debug("threshold=%s, result=%s" % (threshold, best_match))
  18.     return best_match if confidence >= threshold else None

开发者ID:AirtestProject，项目名称:Airtest，代码行数:18，代码来源:template.py


 

示例7: match_dmg_templates

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def match_dmg_templates(self, frame):
  4.         match_mat, max_val, tl = [None]*10, [0]*10, [(0, 0)]*10
  5.         for i in range(0, 10):
  6.             match_mat[i] = cv2.matchTemplate(frame, self.num_img[0],
  7.                 cv2.TM_CCORR_NORMED, mask=self.num_mask[0])
  8.             _, max_val[i], _, tl[i] = cv2.minMaxLoc(match_mat[i])
  9.         # print(max_val[0])
  10.         br = (tl[0][0] + self.num_w, tl[0][1] + self.num_h)
  11.         frame = cv2.rectangle(frame, tl[0], br, (255, 255, 255), 2)
  12.  
  13.         # Multi-template result searching
  14.         # _, max_val_1, _, tl_1 = cv2.minMaxLoc(np.array(match_mat))
  15.         # print(tl_1)
  16.  
  17.  
  18.     # A number of methods corresponding to the various trackbars available.

开发者ID:jpnaterer，项目名称:smashscan，代码行数:18，代码来源:thresholding.py


 

示例8: main

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def main():
  4.     src = cv2.imread('src.jpg', cv2.IMREAD_GRAYSCALE)
  5.     tpl = cv2.imread('tpl.jpg', cv2.IMREAD_GRAYSCALE)
  6.     result = cv2.matchTemplate(src, tpl, cv2.TM_CCOEFF_NORMED)
  7.     result = cv2.normalize(result, dst=None, alpha=0, beta=1,
  8.                            norm_type=cv2.NORM_MINMAX, dtype=-1)
  9.     minVal, maxVal, minLoc, maxLoc = cv2.minMaxLoc(result)
  10.     matchLoc = maxLoc
  11.     draw1 = cv2.rectangle(
  12.         src, matchLoc, (matchLoc[0] + tpl.shape[1], matchLoc[1] + tpl.shape[0]), 0, 2, 8, 0)
  13.     draw2 = cv2.rectangle(
  14.         result, matchLoc, (matchLoc[0] + tpl.shape[1], matchLoc[1] + tpl.shape[0]), 0, 2, 8, 0)
  15.     cv2.imshow('draw1', draw1)
  16.     cv2.imshow('draw2', draw2)
  17.     cv2.waitKey(0)
  18.     print src.shape
  19.     print tpl.shape
  20.     print result.shape
  21.     print matchLoc

开发者ID:cynricfu，项目名称:dual-fisheye-video-stitching，代码行数:21，代码来源:template_matching.py


 

示例9: detect

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def detect(self, z, x):
  4.         k = self.gaussianCorrelation(x, z)
  5.         # 得到响应图
  6.         res = real(fftd(complexMultiplication(self._alphaf, fftd(k)), True))
  7.  
  8.         # pv:响应最大值 pi:相应最大点的索引数组
  9.         _, pv, _, pi = cv2.minMaxLoc(res)
  10.         # 得到响应最大的点索引的float表示
  11.         p = [float(pi[0]), float(pi[1])]
  12.  
  13.         # 使用幅值做差来定位峰值的位置
  14.         if pi[0] > 0 and pi[0] < res.shape[1] - 1:
  15.             p[0] += self.subPixelPeak(res[pi[1], pi[0] - 1], pv, res[pi[1], pi[0] + 1])
  16.         if pi[1] > 0 and pi[1] < res.shape[0] - 1:
  17.             p[1] += self.subPixelPeak(res[pi[1] - 1, pi[0]], pv, res[pi[1] + 1, pi[0]])
  18.  
  19.         # 得出偏离采样中心的位移
  20.         p[0] -= res.shape[1] / 2.
  21.         p[1] -= res.shape[0] / 2.
  22.         
  23.         # 返回偏离采样中心的位移和峰值
  24.         return p, pv
  25.  
  26.     # 基于当前帧更新目标位置

开发者ID:ryanfwy，项目名称:KCF-DSST-py，代码行数:26，代码来源:tracker.py


 

示例10: detect_scale

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def detect_scale(self, image):
  4.         xsf = self.get_scale_sample(image)
  5.  
  6.         # Compute AZ in the paper
  7.         add_temp = cv2.reduce(complexMultiplication(self.sf_num, xsf), 0, cv2.REDUCE_SUM)
  8.  
  9.         # compute the final y
  10.         scale_response = cv2.idft(complexDivisionReal(add_temp, (self.sf_den + self.scale_lambda)), None, cv2.DFT_REAL_OUTPUT)
  11.  
  12.         # Get the max point as the final scaling rate
  13.         # pv:响应最大值 pi:相应最大点的索引数组
  14.         _, pv, _, pi = cv2.minMaxLoc(scale_response)
  15.         
  16.         return pi
  17.  
  18.     # 更新尺度

开发者ID:ryanfwy，项目名称:KCF-DSST-py，代码行数:18，代码来源:tracker.py


 

示例11: imagesearcharea

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def imagesearcharea(image, x1, y1, x2, y2, precision=0.8, im=None):
  4.     if im is None:
  5.         im = region_grabber(region=(x1, y1, x2, y2))
  6.         if is_retina:
  7.             im.thumbnail((round(im.size[0] * 0.5), round(im.size[1] * 0.5)))
  8.         # im.save('testarea.png') usefull for debugging purposes, this will save the captured region as "testarea.png"
  9.  
  10.     img_rgb = np.array(im)
  11.     img_gray = cv2.cvtColor(img_rgb, cv2.COLOR_BGR2GRAY)
  12.     template = cv2.imread(image, 0)
  13.  
  14.     res = cv2.matchTemplate(img_gray, template, cv2.TM_CCOEFF_NORMED)
  15.     min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)
  16.     if max_val < precision:
  17.         return [-1, -1]
  18.     return max_loc

开发者ID:drov0，项目名称:python-imagesearch，代码行数:18，代码来源:imagesearch.py


 

示例12: locate_img

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def locate_img(image, template):
  4.     img = image.copy()
  5.     res = cv2.matchTemplate(img, template, method)
  6.     print res
  7.     print res.shape
  8.     cv2.imwrite('image/shape.png', res)
  9.     min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)
  10.     print cv2.minMaxLoc(res)
  11.     if method in [cv2.TM_SQDIFF, cv2.TM_SQDIFF_NORMED]:
  12.         top_left = min_loc
  13.     else:
  14.         top_left = max_loc
  15.     h, w = template.shape
  16.     bottom_right = (top_left[0] + w, top_left[1]+h)
  17.     cv2.rectangle(img, top_left, bottom_right, 255, 2)
  18.     cv2.imwrite('image/tt.jpg', img)

开发者ID:NetEase，项目名称:airtest，代码行数:18，代码来源:pixelmatch.py


 

示例13: getKeypoints

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def getKeypoints(probMap, threshold=0.1):
  4.  
  5.     mapSmooth = cv2.GaussianBlur(probMap, (3, 3), 0, 0)
  6.     mapMask = np.uint8(mapSmooth>threshold)
  7.     keypoints = []
  8.     contours = None
  9.     try:
  10.         #OpenCV4.x
  11.         contours, _ = cv2.findContours(mapMask, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
  12.     except:
  13.         #OpenCV3.x
  14.         _, contours, _ = cv2.findContours(mapMask, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
  15.  
  16.     for cnt in contours:
  17.         blobMask = np.zeros(mapMask.shape)
  18.         blobMask = cv2.fillConvexPoly(blobMask, cnt, 1)
  19.         maskedProbMap = mapSmooth * blobMask
  20.         _, maxVal, _, maxLoc = cv2.minMaxLoc(maskedProbMap)
  21.         keypoints.append(maxLoc + (probMap[maxLoc[1], maxLoc[0]],))
  22.  
  23.     return keypoints

开发者ID:PINTO0309，项目名称:MobileNetV2-PoseEstimation，代码行数:23，代码来源:openvino-usbcamera-cpu-ncs2-async.py


 

示例14: _locate_target

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def _locate_target(self, score):
  4.         def subpixel_peak(left, center, right):
  5.             divisor = 2 * center - left - right
  6.             if abs(divisor) < 1e-3:
  7.                 return 0
  8.             return 0.5 * (right - left) / divisor
  9.  
  10.         _, _, _, max_loc = cv2.minMaxLoc(score)
  11.         loc = np.float32(max_loc)
  12.  
  13.         if max_loc[0] in range(1, score.shape[1] - 1):
  14.             loc[0] += subpixel_peak(
  15.                 score[max_loc[1], max_loc[0] - 1],
  16.                 score[max_loc[1], max_loc[0]],
  17.                 score[max_loc[1], max_loc[0] + 1])
  18.         if max_loc[1] in range(1, score.shape[0] - 1):
  19.             loc[1] += subpixel_peak(
  20.                 score[max_loc[1] - 1, max_loc[0]],
  21.                 score[max_loc[1], max_loc[0]],
  22.                 score[max_loc[1] + 1, max_loc[0]])
  23.         offset = loc - np.float32(score.shape[1::-1]) / 2
  24.  
  25.         return offset

开发者ID:huanglianghua，项目名称:open-vot，代码行数:25，代码来源:kcf.py


 

示例15: SMAvgLocalMax

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def SMAvgLocalMax(self, src):
  4.         # size
  5.         stepsize = pySaliencyMapDefs.default_step_local
  6.         width = src.shape[1]
  7.         height = src.shape[0]
  8.         # find local maxima
  9.         numlocal = 0
  10.         lmaxmean = 0
  11.         for y in range(0, height-stepsize, stepsize):
  12.             for x in range(0, width-stepsize, stepsize):
  13.                 localimg = src[y:y+stepsize, x:x+stepsize]
  14.                 lmin, lmax, dummy1, dummy2 = cv2.minMaxLoc(localimg)
  15.                 lmaxmean += lmax
  16.                 numlocal += 1
  17.         # averaging over all the local regions
  18.         return lmaxmean / numlocal
  19.     # normalization specific for the saliency map model

开发者ID:tyarkoni，项目名称:pliers，代码行数:19，代码来源:pySaliencyMap.py


 

示例16: matchAB

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def matchAB(fileA, fileB):
  4.     # 读取图像数据
  5.     imgA = cv2.imread(fileA)
  6.     imgB = cv2.imread(fileB)
  7.  
  8.     # 转换成灰色
  9.     grayA = cv2.cvtColor(imgA, cv2.COLOR_BGR2GRAY)
  10.     grayB = cv2.cvtColor(imgB, cv2.COLOR_BGR2GRAY)
  11.  
  12.     # 获取图片A的大小
  13.     height, width = grayA.shape
  14.  
  15.     # 取局部图像，寻找匹配位置
  16.     result_window = np.zeros((height, width), dtype=imgA.dtype)
  17.     for start_y in range(0, height-100, 10):
  18.         for start_x in range(0, width-100, 10):
  19.             window = grayA[start_y:start_y+100, start_x:start_x+100]
  20.             match = cv2.matchTemplate(grayB, window, cv2.TM_CCOEFF_NORMED)
  21.             _, _, _, max_loc = cv2.minMaxLoc(match)
  22.             matched_window = grayB[max_loc[1]:max_loc[1]+100, max_loc[0]:max_loc[0]+100]
  23.             result = cv2.absdiff(window, matched_window)
  24.             result_window[start_y:start_y+100, start_x:start_x+100] = result
  25.  
  26.     plt.imshow(result_window)
  27.     plt.show()

开发者ID:cangyan，项目名称:image-detect，代码行数:27，代码来源:image_detect_02.py


 

示例17: detect

​ 点赞 6 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def detect(self, z, x):
  4.         k = self.gaussianCorrelation(x, z)
  5.         res = real(fftd(complexMultiplication(self._alphaf, fftd(k)), True))
  6.  
  7.         _, pv, _, pi = cv2.minMaxLoc(res)   # pv:float  pi:tuple of int
  8.         p = [float(pi[0]), float(pi[1])]   # cv::Point2f, [x,y]  #[float,float]
  9.  
  10.         if(pi[0]>0 and pi[0]<res.shape[1]-1):
  11.             p[0] += self.subPixelPeak(res[pi[1],pi[0]-1], pv, res[pi[1],pi[0]+1])
  12.         if(pi[1]>0 and pi[1]<res.shape[0]-1):
  13.             p[1] += self.subPixelPeak(res[pi[1]-1,pi[0]], pv, res[pi[1]+1,pi[0]])
  14.  
  15.         p[0] -= res.shape[1] / 2.
  16.         p[1] -= res.shape[0] / 2.
  17.  
  18.         return p, pv

开发者ID:uoip，项目名称:KCFnb，代码行数:18，代码来源:kcftracker.py


 

示例18: image_search

​ 点赞 5 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def image_search(x_start :int, y_start :int, x_end :int, y_end :int,
  4.                      img :str, threshold :int, bmp :image =None) -> Optional[Tuple[int, int]]:
  5.         """Search the screen for the supplied picture.
  6.         
  7.         Returns a tuple with x,y-coordinates, or None if result is below
  8.         the threshold.
  9.         
  10.         Keyword arguments:
  11.         image     -- Filename or path to file that you search for.
  12.         threshold -- The level of fuzziness to use - a perfect match will be
  13.                      close to 1, but probably never 1. In my testing use a
  14.                      value between 0.7-0.95 depending on how strict you wish
  15.                      to be.
  16.         bmp       -- a bitmap from the get_bitmap() function, use this if you're
  17.                      performing multiple different OCR-readings in succession
  18.                      from the same page. This is to avoid to needlessly get the
  19.                      same bitmap multiple times. If a bitmap is not passed, the
  20.                      function will get the bitmap itself. (default None)
  21.         """
  22.         if not bmp: bmp = Inputs.get_bitmap()
  23.         # Bitmaps are created with a 8px border
  24.         search_area = bmp.crop((x_start + 8, y_start + 8,
  25.                                 x_end + 8, y_end + 8))
  26.         search_area = numpy.asarray(search_area)
  27.         search_area = cv2.cvtColor(search_area, cv2.COLOR_RGB2GRAY)
  28.         template = cv2.imread(img, 0)
  29.         res = cv2.matchTemplate(search_area, template, cv2.TM_CCOEFF_NORMED)
  30.         _, max_val, _, max_loc = cv2.minMaxLoc(res)
  31.         if max_val < threshold:
  32.             return None
  33.         
  34.         return max_loc

开发者ID:kujan，项目名称:NGU-scripts，代码行数:34，代码来源:inputs.py


 

示例19: MatchTemplate

​ 点赞 5 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def MatchTemplate(template, target):
  4.   """Returns match score for given template"""
  5.   res = cv2.matchTemplate(target, template, cv2.TM_CCOEFF_NORMED)
  6.   min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)
  7.   return max_val

开发者ID:cfircohen，项目名称:airport，代码行数:7，代码来源:solver.py


 

示例20: match_img

​ 点赞 5 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def match_img(self, image, target, value, rematch=False):
  4.         """
  5.         :param image: 原始图片
  6.         :param target: 匹配模板
  7.         :param value: 匹配阈值
  8.         :param rematch: false,初赛水印,true复赛水印
  9.         :return: 水印外轮廓坐标,原始图片灰度图,水印内轮廓
  10.         """
  11.         img_rgb = cv2.imread(image)
  12.         h, w, c = img_rgb.shape
  13.         img_gray = cv2.cvtColor(img_rgb, cv2.COLOR_BGR2GRAY)
  14.         template = cv2.imread(target, 0)
  15.         th, tw = template.shape
  16.         max_v1 = 0
  17.         if not rematch:
  18.             template = template[16:56, 20:186]
  19.         else:
  20.             template = template[18:107, 19:106]
  21.         res = cv2.matchTemplate(img_gray, template, cv2.TM_CCOEFF_NORMED)
  22.         threshold = value
  23.         min_v, max_v, min_pt, max_pt = cv2.minMaxLoc(res)
  24.         if max_v < threshold:
  25.             return False, False, False
  26.         if not rematch:
  27.             template1 = cv2.imread(self.roi_rematch_img_path, 0)
  28.             template1 = template1[18:107, 19:106]
  29.             res1 = cv2.matchTemplate(img_gray, template, cv2.TM_CCOEFF_NORMED)
  30.             min_v1, max_v1, min_pt1, max_pt1 = cv2.minMaxLoc(res1)
  31.             if max_v < max_v1:  # 避免两种水印匹配重叠的情况
  32.                 return False, False, False
  33.         if not rematch:
  34.             x = 20
  35.             y = 16
  36.         else:
  37.             x = 19
  38.             y = 18
  39.         ori_pt = (min(w - tw - 1, max(max_pt[0] - x, 0)), max(0, min(max_pt[1] - y, h - th - 1)))
  40.         return ori_pt, img_gray, max_pt

开发者ID:Mingtzge，项目名称:2019-CCF-BDCI-OCR-MCZJ-OCR-IdentificationIDElement，代码行数:40，代码来源:watermask_process.py


 

示例21: compare_a_template

​ 点赞 5 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def compare_a_template(img_gray, template):  # 函数返回模板匹配的最大值
  4.     """
  5.     将图片与模板对比，比较相似度
  6.     :param img_gray: 灰度图片
  7.     :param template: 模板图片
  8.     :return: 相似度，介于[0，1]
  9.     """
  10.     #img_gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)         # 转为灰度图
  11.     res = cv2.matchTemplate(img_gray, template, cv2.TM_CCOEFF_NORMED)  # 模板匹配
  12.     _, max_val, _, _ = cv2.minMaxLoc(res)
  13.     return max_val     # 返回的是归一化的相似度的最大值,值位于0-1之间

开发者ID:Mingtzge，项目名称:2019-CCF-BDCI-OCR-MCZJ-OCR-IdentificationIDElement，代码行数:13，代码来源:twist_part.py


 

示例22: match_template1

​ 点赞 5 ​

  1. # 需要导入模块: import cv2 [as 别名]
  2. # 或者: from cv2 import minMaxLoc [as 别名]
  3. def match_template1(template, img, plot=False, method=cv2.TM_SQDIFF_NORMED):
  4.     img = cv2.imread(img, 0).copy()
  5.     template = cv2.imread(template, 0)
  6.     w, h = template.shape[::-1]
  7.     if lib == OPENCV:
  8.         res = cv2.matchTemplate(img, template, method)
  9.         min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)
  10.         if method in [cv2.TM_SQDIFF, cv2.TM_SQDIFF_NORMED]:
  11.             top_left = min_loc
  12.         else:
  13.             top_left = max_loc
  14.     else:
  15.         result = match_template(img, template)
  16.         ij = np.unravel_index(np.argmax(result), result.shape)
  17.         top_left = ij[::-1]
  18.  
  19.     bottom_right = (top_left[0] + w, top_left[1] + h)
  20.  
  21.     if plot:
  22.         cv2.rectangle(img, top_left, bottom_right, 255, 5)
  23.         plt.subplot(121)
  24.         plt.imshow(img)
  25.         plt.title('Detected Point'), plt.xticks([]), plt.yticks([])
  26.         plt.subplot(122)
  27.         plt.imshow(template)
  28.  
  29.         plt.show()
  30.  
  31.     return top_left, bottom_right

开发者ID:tobyqin，项目名称:kog-money，代码行数:31，代码来源:match.py


 

 

注：本文中的cv2.minMaxLoc方法示例整理自Github/MSDocs等源码及文档管理平台，相关代码片段筛选自各路编程大神贡献的开源项目，源码版权归原作者所有，传播和使用请参考对应项目的License；未经允许，请勿转载。

声明：本文内容由网友自发贡献，不代表【wpsshop博客】立场，版权归原作者所有，本站不承担相应法律责任。如您发现有侵权的内容，请联系我们。转载请注明出处：https://www.wpsshop.cn/w/Cpp五条/article/detail/79124
推荐阅读
相关标签

Copyright © 2003-2013 www.wpsshop.cn 版权所有，并保留所有权利。

  

闽ICP备14008679号