Pyhton OpenCV：批量识别图片外层轮廓并截图保存

妖狐艹你老母 2023-06-13 06:07 61阅读 0赞

### Pyhton OpenCV：批量识别图片外层轮廓并截图保存 ###

*  一、背景
 *  二、源代码
 *  三、执行结果

--------------------

# 一、背景 #

![图1][1]  
![图2][2]  
  有2个文件夹，其中文件夹a存在多张如上图片，文件夹b为空。现希望读取文件夹a中所有图片，识别黑色矩形并截图保存至文件夹b中（即去掉黑色矩形之外的白色区域）。

# 二、源代码 #

# 系统：window7
    # 编程工具：PyCharm
    # 编程语言：python
    # 前置条件：已安装好opencv相关环境
    
    # coding:utf-8
    import os
    import cv2
    
    
    # 读取文件夹内的所有文件的路径
    def read_file_name(file_dir):
        list_path =[]  # 文件路径
        list_name = []  # 文件名
        for root, dirs, files in os.walk(file_dir):
            """
            print(root)  # 当前目录路径
            print(dirs)  # 当前路径下所有文件夹
            print(files)  # 当前路径下所有文件
           """
            path = root
            if len(dirs) == 0:  # 该目录下不存在文件夹
                for file in files:
                    path1 = path + '//' + file
                    # print(path1)
                    list_path.append(path1)
                    list_name.append(file)
            else:  # 该目录下存在文件夹
                for dir1 in dirs:
                    path = path + '//' + dir1
                    for file in files:
                        path1 = path + '//' + file
                        # print(path1)
                        list_path.append(path1)
                        list_name.append(file)
        return list_path, list_name
    
    
    # 主函数
    if __name__ == '__main__':
        # 读取所有图片的路径
        path_original = "C://Users//Administrator//Desktop//a"  # 原文件夹
        path_target = "C://Users//Administrator//Desktop//b"  # 目标文件夹
        file_path, file_name = read_file_name(path_original)
        i = 0
        for path in file_path:
            # 读取图片
            img_origin = cv2.imread(path)
            # 灰度化
            img_gray = cv2.cvtColor(img_origin, cv2.COLOR_BGR2GRAY)
            # 二值化
            ret, thresh = cv2.threshold(img_gray, 200, 255, cv2.THRESH_BINARY_INV)  # 识别黑色轮廓
            # 查找轮廓
            contours = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
            # 画出最外层轮廓
            x, y, w, h = cv2.boundingRect(contours[0])
            # img_display = img_origin  # 浅拷贝，修改img_display时，同样也修改img_origin
            img_display = img_origin.copy()  # 深拷贝，修改img_display时不会改变img_origin
            cv2.rectangle(img_display, (x, y), (x + w, y + h), (0, 255, 0), 2)
            # 显示图片
            # cv2.imshow("image", img_display)
            # cv2.waitKey(1000)
            # 保存截图
            path_save = path_target + "//" + file_name[i]
            cv2.imwrite(path_save, img_origin[y:y+h, x:x+w])
            i = i + 1
        print("执行结束。")

# 三、执行结果 #

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy_size_16_color_FFFFFF_t_70_pic_center]  
截图后的图片如下所示。

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy_size_16_color_FFFFFF_t_70_pic_center 1]  
![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy_size_16_color_FFFFFF_t_70_pic_center 2]

[1]: https://img-blog.csdnimg.cn/20191114142033568.png#pic_center
[2]: https://img-blog.csdnimg.cn/20191114142145956.PNG?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy,size_16,color_FFFFFF,t_70#pic_center
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy_size_16_color_FFFFFF_t_70_pic_center]: https://img-blog.csdnimg.cn/20191114143720742.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy,size_16,color_FFFFFF,t_70#pic_center
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy_size_16_color_FFFFFF_t_70_pic_center 1]: https://img-blog.csdnimg.cn/2019111414275547.PNG?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy,size_16,color_FFFFFF,t_70#pic_center
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy_size_16_color_FFFFFF_t_70_pic_center 2]: https://img-blog.csdnimg.cn/2019111414363939.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM0ODAxNjQy,size_16,color_FFFFFF,t_70#pic_center