表格图片的框线识别python

最新推荐文章于 2023-05-14 20:00:53 发布

执着的栋栋

最新推荐文章于 2023-05-14 20:00:53 发布

阅读量3k

点赞数 4

分类专栏：爱好文章标签： python图像

本文链接：https://blog.csdn.net/lidongxx/article/details/97939918

版权

爱好专栏收录该内容

18 篇文章 1 订阅

订阅专栏

	gray = cv.cvtColor(src, cv.COLOR_BGR2GRAY)  # 灰度图片
	ret, binary = cv.threshold(gray, 0, 255, cv.THRESH_BINARY | cv.THRESH_OTSU)  # 全局自适应阈值
	dst = cv.adaptiveThreshold(gray, 255, cv.ADAPTIVE_THRESH_GAUSSIAN_C, cv.THRESH_BINARY, 101, 20)
	kernel1 = cv.getStructuringElement(cv.MORPH_RECT, (1, 30))
	dilated1 = cv.dilate(dst, kernel1)
	eroded1 = cv.erode(dilated1, kernel1)
	thresh = cv.adaptiveThreshold(eroded1, 255, cv.ADAPTIVE_THRESH_GAUSSIAN_C, cv.THRESH_BINARY_INV, 11, 5)
	lines = cv.HoughLinesP(thresh, 1, np.pi / 180, 10, minLineLength=200, maxLineGap=5)

	kernel2 = cv.getStructuringElement(cv.MORPH_RECT, (30, 1))
	dilated2 = cv.dilate(dst, kernel2)
	eroded2 = cv.erode(dilated2, kernel2)
	thresh1 = cv.adaptiveThreshold(eroded2, 255, cv.ADAPTIVE_THRESH_GAUSSIAN_C, cv.THRESH_BINARY_INV, 11, 5)
	lines2 = cv.HoughLinesP(thresh1, 1, np.pi / 180, 10, minLineLength=200, maxLineGap=5)