计算机科学 ›› 2016, Vol. 43 ›› Issue (7): 95-100.doi: 10.11896/j.issn.1002-137X.2016.07.016
• 2015年第二十四届全国多媒体学术会议 • 上一篇 下一篇
卢涛,万永静,杨威
LU Tao, WAN Yong-jing and YANG Wei
摘要: 图像分割是机器视觉中的基础问题,基于阈值的图像分割算法依赖于参数调整,但参数调整容易受到局部最小值的影响且需要耗费大量时间,从而降低了分割算法的质量和效率。为了实现图像分割过程中的自适应阈值选择,提出了一种基于稀疏主成分分析和自适应阈值选择的图像分割算法。该算法首先利用稀疏主成分分析感知图像的噪声水平以自适应去噪,其次通过二维直方图感知图像的主干区域内容以自适应获得全局分割阈值,然后通过移动平均法的局部阈值分割算法对图像进行分割,最后将全局阈值分割和局部阈值分割图像结合,从而获得最佳的分割图像结果。在伯克利数据集上的仿真实验结果表明:相比传统的阈值分割算法,该算法在分割边缘的准确性和对噪声的鲁棒性上具有一定的优势,在主客观上均具有较好的分割效果,基于稀疏主成分分析的自适应阈值选择方法提高了图像的分割质量。
[1] Long Jian-wu,Shen Xuan-jing,Zang Hui,et al.An AdaptiveThresholding Algorithm by Background Estimation in Gaussian Scale Space [J].Acta Automatica Sinica,2014,0(8):1773-1782(in Chinese) 龙建武,申铉京,藏慧,等.基于高斯尺度空间下估计背景的自适应阈值分割算法[J].自动化学报,2014,0(8):1773-1782 [2] Wang Yin-yin,Wang Qiu-guang.A Two-dimensional Thresholding Method Based on the Information of Image Edge[J].Journal of Image and Graphics,2007,2(1):78-81(in Chinese) 王玥玥,王秋光.基于图像边缘信息的2维阈值分割方法[J].中国图象图形学报,2007,2(1):78-81 [3] Liu Ya-kun,Yu Shang-yuan,Luo Si-wei.Threshold Image Segmentation Based on Min-max Cut Algorithm[J].Computer Science,2014,1(1):95-99(in Chinese) 刘雅坤,于双元,罗四维.基于最小最大割算法的阈值分割算法[J].计算机科学,2014,1(1):95-99 [4] Thakur P,Madaan N.A survey of image segmentation tech-niques[J].International Journal of Research in Computer Application Androbotics,2014,2(4):158-165 [5] Liu Jin,Jin Wei-dong.Fast thresholding algorithm of 2D Otsu for low SNR image[J].Application Reasearch of Computers,2013,30(10):3168-3173(in Chinese) 刘金,金炜东.噪声图像的快速二维Otsu阈值分割[J].计算机应用技术,2013,0(10):3168-3173 [6] Ma Wen-ke,Wang Ling,He Hao.Local threshold segmentation algorithm for fingerprint images[J].Computer Enginnering and Applications,2009,5(34):177-179(in Chinese) 马文科,王玲,何浩.一种指纹图像的局部阈值分割算法[J].计算机工程与应用,2009,5(34):177-179 [7] Long Jian-wu,Shen Xuan-jing,Chen Hai-peng.Adaptive mini-mum error thresholdign algorithm[J].Acta Automatic Sinica,2012,8(7):1143-1144(in Chinese) 龙建武,申铉京,陈海鹏.自适应最小误差阈值分割算法[J].自动化学报,2012,8(7):1143-1144 [8] Yao Bo-xu.Image denoising research based on non-local sparse models with low-rand matrix decomposition[D].Shanxi:Xi’an University,2012(in Chinese) 姚波旭.基于低秩矩阵分解的非局部稀疏模型图像去噪方法研究[D].西安:西安电子科技大学,2012 [9] Wu Yue.A sparse principal component analysis method basedon projected gradient[D].Shanghai:Fudan University,2012(in Chinese) 吴悦.一种基于投影梯度的稀疏主成分分析方法[D].上海:复旦大学,2012 [10] Shen H,Huang J Z.Sparse principal component analysis via re-gularized low rank matrix approximation[J].Journal of Muliva-riate Analysis,2008,9(6):1015-1034 [11] Bharath H,Pablo A,Ross G,et al.Simultaneous detection and segmentation[J].Lecture Notes in Computer Science ,2014(7):297-312 [12] Niu Fa-fa,Chen Li,Zhang Yong-xin,et al.Canny edge detection algorithm based on robust principal component analysis[J].Journal of Computer Applications,2014,4(6):1727-1730(in Chinese) 牛发发,陈莉,张永新,等.基于鲁棒主成分分析的Canny边缘检测算法[J].计算机应用技术,2014,4(6):1727-1730 |
No related articles found! |
|