叶是植物的重要特征信息，叶片图像提取是植物器官建模和生鲜植物识别的关键步骤。在植物自动识别和叶建模领域具有重要价值。笔者提出了一种基于主动轮廓技术和细胞神经网络的叶图像提取方法，实践表明基于细胞神经网络的可变模板技术实现了对植物叶片轮廓的灵活控制，同时结合了隐含模型和参数模型的特征，提高了提取的精度和鲁棒性。提取结果表明，采用该算法可以有效提取叶脉络。
Abstract
Leaf was important visual characteristic of plant. The leaf image extraction was a key step of modeling plant organs and living plant recognition, and had important value in the fields of automatic identification of plant and modeling. An efficient leaf image extraction method was proposed by combining active contours with cellular neural networks (CNN) in this paper. The active contours based on CNN provided a high flexibility and control for the contour dynamics. This approach had the advantage of applying a priori knowledge, put similar characteristics from both the implicit and parametric models, to improve the precise and robustness of image extraction. The results showed that the calculation method could be used for effectively extracting leaf vein and an ideal test results was obtained.
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1]Wang D, Kerbyson D J. Realistic image synthesis of plant structures for genetic analysis[J]. Image and Vision Computing, 2001, 19: 517-522.
[2]Wang Z, Chi Z, Feng D. Shape based leaf image retrieval, IEE ProcVis[J]. Image Signal Process, 2003, 150(1): 34-43.
[3]Fu Hong. A twostage approach for leaf vein extraction, IEEE int.Conf[J]. Neural Networks & Signal Processing, 2003, 208-211.
[4]Kass M, Witkin A. Snakes: active contour models[J]. International Journal of Computer Vision, 1988(1): 321-331.
[5]Manh A G, Rabatel G, Assemat L, et al. Weed leaf image segmentation by deformable templates J[J]. Agric Engng Res, 2001, 80(2): 139-146.
[6]McInerney T, Terzopoulos D. A dynamic finite element surface model for segmentation and tracking in multidimensional medical images with application to cardiac 4D image analysis[J]. Journal of Computers on Medical Imaging and Graphics, 1995, 19(1): 69-83.
[7]Vilarino D L, Cabello D, Pardo X M, et al. Cellular neural networks and active contours: a tool for image segmentation[J]. Image and Vision Computing, 2003, 21: 189-204.
[8]Chua L O, Yang L. Cellular neural networks: theory[J]. IEEE Transactions on Circuits and Systems, 1988, 35: 1257-1273.
[9]Zhu Y, Yan H. Computerized tumor boundary detection using a Hopfield neural network[J]. IEEE Transactions on Medical Imaging, 1997, 16(1): 55-67.
[10]Caselles V, Catte F, Dibos F. A geometric model for active contours in image processing[J]. Numerical Maths, 1993, 66: 1-31.
[11]Vilarino D L. Discretetime CNN for image segmentation by active contours[J]. Pattern Recognition Letters, 1998, 19(8): 721-734.
[12]Mark S Hammel, Przemyslaw Prusinkiewicz. Modelling Compound Leaves Using Implicit Contours[R]. Proceedings of Computer Graphics International. Japan: Tokyo University, 1992.
基金
收稿日期：2008-04-12修回日期：2008-12-26基金项目：重庆市自然科学基金资助项目(CSTC2006bb2229)；国家自然科学基金资助项目(60773082)；中国博士后科学基金(20080430744,20080430740)作者简介：李云峰(1975—)，讲师，研究方向为计算机信息科学。Email: lyf129@126.com。引文格式：李云峰,曹渝昆,朱庆生. 基于主动轮廓技术的植物叶图像提取方法[J]. 南京林业大学学报:自然科学版,2009,33(3):146-150.
PDF(970230 KB)
