基于视觉皮层机制的轮廓整合的神经网络模型

本文以生物视觉功能及其机理有关的研究成果为主要依据,通过分析、建模和模拟计算等方式建立进化的图像目标轮廓组织整合的理论、方法和模型。首先通过分析初级视觉通道内（LGN—〉视觉皮层V1区—〉视觉皮层V2区神经元）目标轮廓的组织整合机理,建立具有广泛视觉心理学和生理学基础人工神经网络模型;然后将该模型应用于自然场景图像的目标轮廓组织及整合的模拟计算中,并根据实验结果对模型做出分析与评估。

The neural model of contour integration

This paper proposes a neural model of contour integration based on the achievements of the function and mechanism of animal＇s visual. First, by analyzing the mechanism of contour integration in the primary visual route （LGN-〉visual cortical areas VI-〉 visual cortical areas V2）, build a neural model. Second, we will use this model to simulate contour integration in the real world. Afterwards analyzing and evaluate this neural model based on experiment results.