摘要
旨在研究复杂背景下叶片病斑的分割。由于复杂背景会带来巨大的噪声,产生过多的边缘和灰度值不均匀的区域,很容易导致过分割的现象,因此在复杂背景下,很难通过1次分割就完成对叶片病斑的分割。为了解决复杂背景下过分割的现象,提出两步分割的策略。第1步先用笔者提出的各向异性扩散测地线活动轮廓模型(anisotropic diffusion geodesic active contour model,简称AD-GAC模型)进行预分割,在此过程中构造新的边缘检测函数(edge stop function,简称ESF);第2步通过最大熵阈值法完成最终的分割。随后,提取并计算预分割部分各像素灰度值的最大熵,以得到病斑部分与叶片部分的灰度值阈值,通过阈值来完成最后1步的分割。通过MATLAB仿真,可以证明该算法可以有效地将病斑从复杂背景下的叶片上分割出来。研究结果后续的病斑识别作了铺垫。
引文
[1]Mai X C,Meng M Q H,Automatic lesion segmentation from rice leaf blast field images based on random forest[C].RCAR,2016:255-259.
[2]Hu Q X,Tian J,He D J.Wheat leaf lesion color image segmentation with improved multichannel selection based on the ChanVese model[J].Computers and Electronics in Agriculture,2017,135(C):260-268.
[3]Zaitoun N M,Aqel M J.Survey on image segmentation techniques[J].Procedia Computer Science,2015,65:797-806.
[4]Pratondo A,Ong S H,Chui C K,Region growing for medical image segmentation using a modified multiple-seed approach on a multi-core CPU computer[C]//The 15th International Conference on Biomedical Engineering,2013:112-115.
[5]Vincent L,Soille P,Watersheds in digital spaces:an efficient algorithm based on immersion simulations[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1991,13(6):583-598.
[6]Zhou S P,Wang J J,Zhang S,et al.Active contour model based on local and global intensity information for medical image segmentation[J].Neurocomputing,2016,186:107-118.
[7]Kass M,Witkin A,Terzopoulos D.Snakes:active contour models[J].International Journal of Computer Vision,1988,1(4):321-331.
[8]Zeng L,Chen J,Xu Y F,et al.Level set method for image segmentation based on local variance and improved intensity inhomogeneity model[J].IET Image Processing,2016,10(12):1007.
[9]Csaelles V,Kimmel R,Sapiro G.Geodesic active contours[C]//Proceedings of IEEE International Conference on Computer Vision,1995:694-699.
[10]Perona P,Malik J.Scale-space and edge detection using anisotropic diffusion[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1990,12(7):629-639.
[11]Caselles V,Catt F,Coll T,et al.A geometric model for active contours in image processing[J].Numerische Mathematik,1993,66(1):1-31.
[12]Osher S,Fedkiw R.Level set methods and dynamic implicit surfaces[M].New York:Springer-Verlag,2002:158-174.
[13]Li C M,Xu C Y,Gui C F,et al.Level set evolution without re-initialization:a new variational formulation[C]//IEEE Conf Comput Vis Pattern Recognit,2005:430-436.
[14]Long J W,Zhang J X,Xiang N,et al.An iterative maximum entropy thresholding algorithm[C]//2016 International Conference on Cyberworlds,2016:171-174.
[15]Tian J,Hu Q X,Ma X Y.Color image segmentation of plant lesion using improved C-V model based on Gaussian distribution[J].Transactions of the Chinese Society of Agricultural Engineering,2013,29(16):166-173.