基于视觉注意机制的图像压缩感知复原

Image compression sensing restoration based on visual attention mechanism

对于低采样、低质航运图像的复原问题,提出了一种新的图像压缩感知复原算法。首先对图像进行预处理,然后由人眼视觉注意机制模型确定注意点位置,并采用自适应遗传算法对注意点像素值进行寻优,最后复原获得复原图像。实验结果表明,所提出的基于视觉注意机制的图像压缩感知复原算法不仅提高了图像的复原质量,同时很好地保存了图像的细节信息。

For the low-sampling and low-quality navigation images restoration problem,this paper proposed a new image compression perception recovery algorithm. Firstly,the image is preprocessed. Then the attention point is determined by the human visual attention mechanism model,followed by using the adaptive genetic algorithm( GA) to find the pixel values of the attention point. Finally,a restored image is obtained. The experimental results show that the image compressed sensing recovery algorithm based on visual attention mechanism proposed in this paper,not only can improve the quality of reconstructed image,but also can better preserve the detail information of image.