基于改进的Scharr算法的海上舰船图像边缘检测
|
摘要
为了提高海上舰船图像边缘检测的准确性,针对传统的边缘检测方法存在检测效果不理想、抗噪能力低等情况,现提出一种改进的Scharr边缘检测算法。该算法首先通过统计滤波去噪预处理方法,降低图像中噪声的不良影响;然后采用0°,45°,90°和135°四个方向的Scharr应双阈值选取算法确定的最大和最小阈值进行边缘检测,提高了边缘的完整性和连续性,边缘图像质量较好。实验结果表明该算法提取的边缘图像比传统的边缘检测算法具有更好的边缘检测效果且抗噪能力明显增强。
In view of the obvious shortage of traditional edge detection methods,which is low edge detection quality and cannot segment accurately,this paper proposes an improved Scharr edge detection algorithm to improve accuracy in edge detection in marine ship image. Firstly,the statistical filtering algorithm is used to decrease the influence affected by noise. This paper uses four Scharr operator templates of 0°,45°,90° and 135° different directions to detect marine ship edges better and enhances the directionality of edge detection. The maximum threshold and the mimimun threshold are selected by the improved adaptive double threshold selection algorithm,which is used to detect ship edges,which improves integrity,continuity and edge image quality. The experimental results show that this method is better than the traditional edge detection,and has better edge detection quality and anti-noise property.
In view of the obvious shortage of traditional edge detection methods,which is low edge detection quality and cannot segment accurately,this paper proposes an improved Scharr edge detection algorithm to improve accuracy in edge detection in marine ship image. Firstly,the statistical filtering algorithm is used to decrease the influence affected by noise. This paper uses four Scharr operator templates of 0°,45°,90° and 135° different directions to detect marine ship edges better and enhances the directionality of edge detection. The maximum threshold and the mimimun threshold are selected by the improved adaptive double threshold selection algorithm,which is used to detect ship edges,which improves integrity,continuity and edge image quality. The experimental results show that this method is better than the traditional edge detection,and has better edge detection quality and anti-noise property.
引文
[1]刘亭利.模糊增强的Canny算法在船舶红外图像分割中的应用[J].舰船科学技术,2017,39(04):184-186.
[2]闫凤,王瑞星,任立胜.小波变换在舰船边缘检测及尺寸测量中的应用[J].舰船科学技术,2016,38(10):70-72.
[3]李森.小波变换在舰船图像尺寸测量方法中的应用[J].舰船科学技术,2015,37(03):177-180.
[4]何泊.全变分型图像去噪与去模糊问题[D].开封:河南大学,2014:20-25.
[5]段军,高翔.基于统计滤波的自适应双阈值改进canny算子边缘检测算法[J].激光杂志,2015,36(01):10-12.
[6]杨廷草.基于FPGA的PCBA图像边缘检测系统设计[D].苏州:苏州大学,2016:14-16.
[7]阳升.手指静脉特征提取算法的研究[D].长沙:湖南大学,2015:23-27.
[8]沈德海,侯建,鄂旭等.基于Sobel的多方向算子模板边缘检测算法[J].现代电子技术,2015,38(04):91-93,97.
[9]周明.相干激光雷达强度像边缘检测算法及性能评价仿真研究[D].哈尔滨:哈尔滨工业大学,2011:10-14.
[10]高翔.医学细胞显微图像预处理及分割技术的研究[D].包头:内蒙古科技大学,2016:18-22.
[11]张志顺,奚建清,刘勇.基于改进Canny算子的CT图像边缘检测[J].微电子学与计算机,2013,30(09):9-12.
[12]盛会平,林王鹏,周勇.信号处理在舰船目标识别中研究发展综述[J].舰船电子工程,2013,33(12):3-6,111.
[13]周胜,常君明,李洞,陈曦,吴冕,甘祥麟.基于计算机视觉的贴片元件定位检测算法研究[J].舰船电子工程,2017,37(09):114-116.
[2]闫凤,王瑞星,任立胜.小波变换在舰船边缘检测及尺寸测量中的应用[J].舰船科学技术,2016,38(10):70-72.
[3]李森.小波变换在舰船图像尺寸测量方法中的应用[J].舰船科学技术,2015,37(03):177-180.
[4]何泊.全变分型图像去噪与去模糊问题[D].开封:河南大学,2014:20-25.
[5]段军,高翔.基于统计滤波的自适应双阈值改进canny算子边缘检测算法[J].激光杂志,2015,36(01):10-12.
[6]杨廷草.基于FPGA的PCBA图像边缘检测系统设计[D].苏州:苏州大学,2016:14-16.
[7]阳升.手指静脉特征提取算法的研究[D].长沙:湖南大学,2015:23-27.
[8]沈德海,侯建,鄂旭等.基于Sobel的多方向算子模板边缘检测算法[J].现代电子技术,2015,38(04):91-93,97.
[9]周明.相干激光雷达强度像边缘检测算法及性能评价仿真研究[D].哈尔滨:哈尔滨工业大学,2011:10-14.
[10]高翔.医学细胞显微图像预处理及分割技术的研究[D].包头:内蒙古科技大学,2016:18-22.
[11]张志顺,奚建清,刘勇.基于改进Canny算子的CT图像边缘检测[J].微电子学与计算机,2013,30(09):9-12.
[12]盛会平,林王鹏,周勇.信号处理在舰船目标识别中研究发展综述[J].舰船电子工程,2013,33(12):3-6,111.
[13]周胜,常君明,李洞,陈曦,吴冕,甘祥麟.基于计算机视觉的贴片元件定位检测算法研究[J].舰船电子工程,2017,37(09):114-116.