基于角点检测的二维图像边缘平直化算法研究
|
摘要
为了解决二维图像中物体的粗糙直边问题,需要对图像进行平直化处理和近似直线的曲边缘进行直线化处理。文中首先对二值化后的图像利用八邻域轮廓跟踪算法获取按一定顺序排列二维图像边缘的坐标集合。然后使用Harris角点检测法得出图像的角点集合。再求取两集合的交集得出按一定顺序排列的角坐标。进而计算起始角点与结束角点连线与起始角点和两角点间轮廓的中点的连线的夹角是否小于给定值来判断为原图像两角点的连线是否直线。最后利用"两点一线"原理把二维图像边缘平直化。通过大量实验证明,该算法对于二维图像边缘平直化不仅有着非常好的效果,而且处理速度快,具有实时性、鲁棒性和实用性。
In order to solve the problem of rough straight edges of objects in two-dimensional images, it is necessary to straighten the image and straighten the curved edges of the approximate line. In this paper, the binarized image is used to obtain the coordinate sets of the edges of the two-dimensional image in a certain order by using the 8-neighbor contour tracking algorithm. The corner point set of the image is then derived using Harris corner detection. Then the intersection of the two sets is calculated to get the angular coordinates arranged in a certain order. Further, it is determined whether the angle between the line connecting the starting corner point and the ending corner point and the line connecting the starting point and the midpoints of the contours between the two corner points is smaller than a given value to determine whether the line connecting the two corner points of the original image is straight. Finally, the "two points and one line" principle is used to straighten the edges of the two-dimensional image. Through a large number of experiments, the proposed algorithm not only has a very good effect on the straightening of the edge of the two-dimensional image, but also has a fast processing speed, real-time, robustness and practicability.
In order to solve the problem of rough straight edges of objects in two-dimensional images, it is necessary to straighten the image and straighten the curved edges of the approximate line. In this paper, the binarized image is used to obtain the coordinate sets of the edges of the two-dimensional image in a certain order by using the 8-neighbor contour tracking algorithm. The corner point set of the image is then derived using Harris corner detection. Then the intersection of the two sets is calculated to get the angular coordinates arranged in a certain order. Further, it is determined whether the angle between the line connecting the starting corner point and the ending corner point and the line connecting the starting point and the midpoints of the contours between the two corner points is smaller than a given value to determine whether the line connecting the two corner points of the original image is straight. Finally, the "two points and one line" principle is used to straighten the edges of the two-dimensional image. Through a large number of experiments, the proposed algorithm not only has a very good effect on the straightening of the edge of the two-dimensional image, but also has a fast processing speed, real-time, robustness and practicability.
引文
[1]李雪.图像配准中角点检测算法的研究与比较[J].信息化建设,2016(3):263.
[2]高建林.基于边缘的角点分类和描述算法[J].电子科技,2016(12):81-84.
[3] MOKHTARIAN F, MOHANNA F. Performance Evaluation of Corner Detectors using Consistency and Accuracy Measures[J].Computer Vision and Image Understanding,2006,102(1):81-94.
[4]丁雄飞,张春燕.基于Moravec算子和改进的SIFT算法的图像匹配[J].合肥学院学报(自然科学版),2013(3):40-42.
[5] KITCHEN L, ROSENFELD A. Analysis grey level comer detection[J].Pattern Recognition Leters,1999(6):81-87.
[6]曾接贤,李炜烨.曲率尺度空间与链码方向统计的角点检测[J].中国图象图形学报,2014(2):234-242.
[7]吴海滨,周英蔚,周雨润,等.基于Hessian矩阵的黑白棋盘格角点检测[J].大气与环境光学学报,2013(5):395-400.
[8]林鹏岳,李玲玲,李翠华.一种改进的快速SUSAN角点检测算法[J].计算机与现代化,2010(2):68-72.
[9]何春.一种基于直方图的图像二值化算法[J].宜宾学院学报,2016(12):53-55.
[10] WANG Y T, CHEN Y Q, LI J.The Harris corner detection method based on three scale invariance spaces[J]. International Journal of Computer Science Issues, 2012, 9(6):18-22.
[11]余博,郭雷,赵天云.Freeman链码描述的曲线匹配方法[J].西北工业大学自动化学院,2012(4):5-8.
[12]娄联堂,韦茜妤,陈佳骐,等.基于图像连续表示的角点检测[J].中南民族大学学报(自然科学版),2016(2):151-156.
[13] TSAI Y H, CHUNG K L. Region-filling algorithm on bin-code based contour and its implementation[J]. Computer Graph., 2000,24(4):529-537.
[14] CHANG F, CHEN C J. A Component-Labeling Algorithm Using Contour Tracing Technique[J]. Computer Vision&Image Under standing, 2004, 93(2):206-220.
[15]黄坚,刘桂雄,林镇秋.基于多角点结合的机箱标准件图像特征提取方法[J].中国测试,2017(9):123-127.
[2]高建林.基于边缘的角点分类和描述算法[J].电子科技,2016(12):81-84.
[3] MOKHTARIAN F, MOHANNA F. Performance Evaluation of Corner Detectors using Consistency and Accuracy Measures[J].Computer Vision and Image Understanding,2006,102(1):81-94.
[4]丁雄飞,张春燕.基于Moravec算子和改进的SIFT算法的图像匹配[J].合肥学院学报(自然科学版),2013(3):40-42.
[5] KITCHEN L, ROSENFELD A. Analysis grey level comer detection[J].Pattern Recognition Leters,1999(6):81-87.
[6]曾接贤,李炜烨.曲率尺度空间与链码方向统计的角点检测[J].中国图象图形学报,2014(2):234-242.
[7]吴海滨,周英蔚,周雨润,等.基于Hessian矩阵的黑白棋盘格角点检测[J].大气与环境光学学报,2013(5):395-400.
[8]林鹏岳,李玲玲,李翠华.一种改进的快速SUSAN角点检测算法[J].计算机与现代化,2010(2):68-72.
[9]何春.一种基于直方图的图像二值化算法[J].宜宾学院学报,2016(12):53-55.
[10] WANG Y T, CHEN Y Q, LI J.The Harris corner detection method based on three scale invariance spaces[J]. International Journal of Computer Science Issues, 2012, 9(6):18-22.
[11]余博,郭雷,赵天云.Freeman链码描述的曲线匹配方法[J].西北工业大学自动化学院,2012(4):5-8.
[12]娄联堂,韦茜妤,陈佳骐,等.基于图像连续表示的角点检测[J].中南民族大学学报(自然科学版),2016(2):151-156.
[13] TSAI Y H, CHUNG K L. Region-filling algorithm on bin-code based contour and its implementation[J]. Computer Graph., 2000,24(4):529-537.
[14] CHANG F, CHEN C J. A Component-Labeling Algorithm Using Contour Tracing Technique[J]. Computer Vision&Image Under standing, 2004, 93(2):206-220.
[15]黄坚,刘桂雄,林镇秋.基于多角点结合的机箱标准件图像特征提取方法[J].中国测试,2017(9):123-127.