杭州电子科技大学自动化学院;
目的 基于初级视觉皮层非经典感受野亚区外周抑制特性机理,提出一种新型轮廓检测方法。方法 模拟初级视皮层经典感受野受到外部刺激的响应特性,构建多方位二维Gabor滤波器模型,实现初级轮廓的提取。之后基于初级皮层非经典感受野亚区的结构特性,提出非经典感受野亚区侧抑制模型,实现纹理抑制。对于视觉信息传递的前馈机制,采用二维Gaussian函数模拟神经节细胞对信息的处理,信息跨层级传递,提高响应速率。模拟人眼捕捉全局信息的特性,进行轮廓修正,最终得到轮廓图。结果 针对RUG40图像库的测试,与其他轮廓检测算法进行定性和定量分析比较;BSDS500中任意200张图像的平均准确度(AP)达0.703。结论 新型轮廓检测方法能更有效突显主体轮廓并且抑制纹理背景。
12 | 0 | 13 |
下载次数 | 被引频次 | 阅读次数 |
[1] Zhong H, Wang R. A visual-degradation-inspired model with HSV color-encoding for contour detection[J]. Journal of Neuroscience Methods, 2022, 369:109423.
[2] Fang T, Fan Y, Wu W. Salient contour detection on the basis of the mechanism of bilateral asymmetric receptive fields[J]. Signal, Image and Video Processing, 2020, 14:1461-1469.
[3] Lin C, Zhang Q, Cao Y. Multi-scale contour detection model based on fixational eye movement mechanism[J]. Signal, Image and Video Processing, 2020, 14:57-65.
[4]李健,范影乐.基于信息流多级结构响应的轮廓检测模型[J].传感技术学报,2024,37(2):288-296.
[5] Chen Z, Cai R. Contour detection by simulating the curvature cell in the visual cortex and its application to object classification[J]. IEEE Access, 2020, 8:74472-74484.
[6] Zhang Q, Lin C, Li F. Application of binocular disparity and receptive field dynamics:A biologically-inspired model for contour detection[J]. Pattern Recognition, 2021, 110:107657.
[7]武薇,韩显修,范影乐.基于双视通路交互感知的轮廓检测方法[J].电子与信息学报,2022,44(7):2512-2521.
[8]杨鑫,朱恒亮,毛国君.双特征流融合和边界感知的显著性目标检测[J].计算机工程与应用,2024,60(10):227-236.
[9]张贞光.基于视觉通路生理机制的轮廓检测模型[D].柳州:广西科技大学,2023.
[10] Wei M, Chaofan G, Shibiao X, et al. Top-down guided fusion network for object contour detection[J]. Journal of Computer-Aided Design&Computer Graphics, 2021, 33(6):855-863.
[11]杨谦,齐翔林,汪云九.简单细胞方位选择性感受野组织形成的神经网络模型[J].中国科学:C辑,2000,30(4):412-420.
[12] Hubel DH, Wiesel TN. Receptive fields of single neurones in the cat's striate cortex[J]. The Journal of Physiology, 1959, 148(3):574.
[13] Jones HE, Grieve KL, Wang W, et al. Surround suppression in primate V1[J]. Journal of Neurophysiology, 2001, 86(4):2011-2028.
[14] Gilbert C D, Li W. Top-down influences on visual processing[J]. Nature Reviews Neuroscience, 2013, 14(5):350-363.
[15]张明琦,范影乐,武薇.基于初级视通路视觉感知机制的轮廓检测方法[J].自动化学报,2020,46(2):264-273.
[16]杨开富.前端视觉通路信息加工的计算模型及应用研究[D].成都:电子科技大学,2016.
[17] Pingping C, Zhefei C, Yingle F, et al. Image contour detection based on visual pathway information transfer mechanism[J]. Neural Processing Letters,2024,56(1):825-829.
[18]杜仕荣,范影乐,蔡哲飞,等.结合双视通路与尺度信息融合的轮廓检测方法[J].中国图象图形学报,2024.
[19] Lin C, Wen Z Q, Xu G L, et al. A bio-inspired contour detection model using multiple cues inhibition in primary visual cortex[J].Multimedia Tools and Applications, 2022, 81(8):11027-11048.
[20] Elharrouss O, Hmamouche Y, Idrissi A K, et al. Refined edge detection with cascaded and high-resolution convolutional network[J].Pattern Recognition, 2023, 138:109361.
基本信息:
DOI:10.16289/j.cnki.1002-0837.2025.01011
中图分类号:TP391.41
引用信息:
[1]章敬艳,范影乐,房涛.基于非经典感受野亚区的轮廓检测方法[J].航天医学与医学工程,2025,36(01):65-68+74.DOI:10.16289/j.cnki.1002-0837.2025.01011.
基金信息: