摘要
边缘提取是图像处理的基础工作,如何精确、有效地提取边缘是图像处理领域相关学者讨论的热点问题,由此产生的各种边缘检测方法层出不穷并且得到了很好的应用,但这些方法都无法达到人眼识别物体边缘的精确程度。目前脉冲耦合神经网络(pulse coupled neural network,PCNN)是图像处理领域较为接近生物视觉进行图像处理的有力工具。改进基本的PCNN模型,提出了一种新的模拟生物视觉提取图像边缘的方法,该改进方法有效地利用了PCNN的特性。将该方法应用于医学图像的边缘提取,并与几种经典边缘检测算法、基本的PCNN方法进行比较,通过实验结果证明改进的方法提取的边缘更加完整、清晰,并且对椒盐噪声具有较强的抑制能力。
Edge extraction is foundation work of image processing,how accurate,efficient extraction edge image processing area is related to discuss the issues,scholars resulting from various edge detection method to emerge in endlessly and got very good application.But these methods are unable to reach human recognition on the edge of the object precision degree.Current pulse coupled neural network is relatively close image processing fields of biological visual image processing is a powerful tool.This improvement of basic PCNN model,put forward a new simulation biological visual image edge extraction method,the method of improving the effective used of the characteristics of the PCNN.It applied the improved method applied to the medical image edge extraction,and edge detection algorithm with several classic PCNN and basic methods.Through experiment results prove the edge extraction method is more complete,clear and salt pepper noise with strong inhibitory ability.
出处
《计算机应用研究》
CSCD
北大核心
2010年第11期4389-4393,共5页
Application Research of Computers
基金
国家自然科学基金资助项目(60970098
60803024)
国家自然科学基金重大研究计划资助项目(90715043)
国家教育部高等学校博士点基金资助项目(20090162110055)
新教师基金资助项目(200805331107)
浙江大学计算机辅助设计与图形学国家重点实验室开发课题(A1011
A0911)
湖南省教育厅科研资助项目(09C745)
