opencv matchShapes()

最新推荐文章于 2025-05-04 15:10:03 发布

qq76211822

最新推荐文章于 2025-05-04 15:10:03 发布

阅读量4.4k

点赞数

CC 4.0 BY-SA版权

分类专栏： OPENCV

本文链接：https://blog.youkuaiyun.com/sz76211822/article/details/72379817

OPENCV 专栏收录该内容

40 篇文章

订阅专栏

本文介绍了一种快速的图像处理技术，通过使用Canny算法提取图像边缘，然后利用findContours找到图像中的轮廓。接着通过matchShapes函数来比较这些轮廓与模板轮廓的相似度，从而实现目标检测。这种方法适用于能够获取到单一完整轮廓的情况。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

First you will need to get contours of shapes from image. For example use Canny() to find edges in images. Than use findContours() on image of edges to get contours.

Than you can use matchShapes() to find contour that is most similar to your model contour (among all contours extracted by findContours()).

Note that this approach is very fast comparing to other methods of object detection, but if you can't get single contour around your book you will not able to find it. matchShapes() compares single contour to single contour.