img_matrix = []
# 标题矩阵
title_matrix = []
# 对每一张图片进行处理
for path in img_path:
img_list = []
title_list = []
img = cv2.imread(path)
img_list.append(img)
filename = filecm.short_name(path)
title_list.append('原图-%s' % filename)
# 灰度转换
gray = opencvcm.get_gray(img, tmp_path, filename, img_list, title_list)
# 平滑转
blur = opencvcm.get_blur(gray, True, (3, 3), 5, tmp_path, filename, img_list, title_list)
# 边缘检测
sobel = opencvcm.get_sobel(blur, 1, 0, 3, cv2.CV_8U, tmp_path, filename, img_list, title_list)
# 二值化
thresh = opencvcm.get_thresh(sobel, 170, cv2.THRESH_BINARY, tmp_path, filename, img_list, title_list)
# 边界
kernel_dilate = cv2.getStructuringElement(cv2.MORPH_RECT, (9, 7))
kernel_erode = cv2.getStructuringElement(cv2.MORPH_RECT, (9, 1))
edges = opencvcm.get_edges(thresh, 'DED-DE', kernel_dilate, kernel_erode, tmp_path, filename, img_list,
title_list)
# 查找车牌区域，指定最小面积，宽高比范围
rois = opencvcm.find_rois(edges, min_area=2000, min_wh_ratio=2, max_wh_ratio=5)
img_plate = None
filecm.makedir(tmp_path)
# KNN算法对象（初始化并训练好）
knn = cvknncm.get_hand_digits_knn()
# 载入测试图片
img_test = cv2.imread('./images/cv_img_number2_test.jpg')
# 图片列表
img_list = []
# 标题列表
title_list = []
img_list.append(img_test)
title_list.append('原图+结果')
gray = opencvcm.get_gray(img_test, tmp_path, "", img_list, title_list)
# 读取图片边界
edges = opencvcm.get_edges(gray, 'D-E', None, None, tmp_path, "", img_list,
title_list)
# 阈值处理
thresh = opencvcm.get_thresh(edges, 21, cv2.THRESH_BINARY, tmp_path, "",
img_list, title_list)
# 查找兴趣点区域
rois = opencvcm.find_rois(thresh)
digits = []
id = 0
for roi in rois:
(x, y, w, h) = roi
id += 1
# 查询兴趣区域对应的数字
digit, th = cvknncm.find_hand_digit_knn(knn,