一种带有连续变化分辨率的中央凹视觉处理方法研究

Research of a method for the foveated visual processing with continuous resolution variable

针对以往中央凹视觉处理过程中,中央凹区域太小,分辨率下降过快,难以满足模式识别、跟踪等问题。基于可分解映射,提出一种带有连续变化分辨率的中央凹视觉处理方法。实验结果表明:在保持中央凹图像数据量不变的基础上,该方法可以提供更大的中央凹区域。

In view of the previous process of foveated visual processing, fovea region being too small,resolution falling too fast,it’s difficult to meet the problem of pattern recognition,tracking and other applications. Based on separable mappings, this paper proposes a method for the foveated visual processing with continuous resolution variable. The experimental results show that in the fovea image data size unchanged, on the basis of this method can provide greater foveal region.