基于点阵神经元响应时空信息的菌落图像边缘检测

Edge Detection in Colony Images Based on Spatial and Temporal Information of Responses from Lattice Neurons

目的基于图像边缘检测所获得的目标轮廓细节质量对于后续图像分析或理解过程具有的重要作用,提出一种基于视觉机制的图像边缘检测新方法。方法构造视觉感受野窗口在经过Log-Gabor滤波器重构后的图像上滑动,对感受野内点阵神经元的发放时刻进行次序编码;同时考虑神经元之间的侧向抑制作用,最后经灰度逆映射获取边缘图像。结果基于点阵神经元响应时空信息的图像强弱边缘检测方法能够有效提取图像的边缘信息,同时也能够体现更多的图像细节。结论本文研究的图像边缘检测方法为利用视觉生理特性进行图像处理提供了崭新而有效的思路。

Objective Details of objects＇ contour can be acquired from edge detection, and the details are con- sidered important for later process of image analysis or understanding, so a new method of image edge detection based on vision mechanism was proposed in this paper. Methods A visual receptive field＇s window was constructed to slither on the image which has been reconstructed by Log-Gabor filter, and then the rank order of fire times of lattice neurons in the receptive field was encoded. Meanwhile, the lateral inhibition between neurons was considered. In the end, the edge image was acquired by gray inverse mapping. Results The meth- od of edge detection based on spatial and temporal information of responses from the lattice neurons extracted edge information effectively and reflected more image details. Conclusion The method of edge detection dis- cussed in the paper provides a brand-new and effective idea for image processing based on visual physiological characteristics.