仿视皮层机制的随机点视频序列运动特征提取

发布时间：2018-04-28 12:31

本文选题：中颞叶区 + 运动特征　；参考：《电子科技大学学报》2017年04期

【摘要】：为了探索视皮层对多方向刺激的运动信息处理机制,依据生物实验中部分中颞叶区(MT)细胞对随机点视频刺激运动感知的侧偏特性,提出了一种仿视皮层机制的随机点视频序列运动特征提取模型。首先,采用Von Mises函数拟合初级视皮层(V1)简单细胞的感受野,再结合非线性的能量模型、调谐和非调谐正则化方法等模拟实现V1复杂细胞对运动信息的方向感知;其次,提出了级联前馈方法对V1复杂细胞响应进行线性加权求和,得到MT细胞的响应输出;最后,利用生物实测MT细胞响应数据对新模型的有效性进行了实验验证。仿真结果表明模型的输出结果与恒河猴的运动感知实验数据基本吻合,能够较好地模拟视皮层对不同夹角的多方向随机点视频序列的运动感知机制,为提取复杂运动序列的特征提供新的方法,进而为视觉类脑计算奠定基础。
[Abstract]:In order to explore the mechanism of motor information processing in the visual cortex for multi-directional stimuli, according to the side bias characteristics of MTcells in the middle temporal lobe region of some biological experiments, the perception of motion of random video stimulation was observed. This paper presents a motion feature extraction model of random spot video sequences based on the mechanism of visual cortex. First, the Von Mises function is used to fit the receptive field of the primary visual cortex (V1) simple cells, and then the nonlinear energy model, tunable and untuned regularization methods are used to realize the directional perception of the motion information of the V1 complex cells. A cascaded feedforward method is proposed to sum the responses of complex V1 cells linearly, and the response output of MT cells is obtained. Finally, the validity of the new model is verified by using biological measured MT cell response data. The simulation results show that the output of the model is basically consistent with the experimental data of motion perception of rhesus monkey, and it can well simulate the motion perception mechanism of the visual cortex to the multi-directional random point video sequences with different angles. It provides a new method for extracting the features of complex motion sequences and lays the foundation for visual brain-like computing.
【作者单位】：空军工程大学航空航天工程学院;
【基金】：国家自然科学基金(61372167,61379104)
【分类号】：TP391.41

【相似文献】