基于区域选择的红外弱小目标超分辨率复原算法

红外成像技术以其诸多优点成为智能化光电探测方面的主流研究方向,然而,红外弱小目标图像却有细节特征少、信噪比低等特点,因此考虑到使用超分辨率复原算法对其进行复原,为图像提供更多的细节信息。本文分析了凸集投影法的基本原理,针对其运行时间长的特点,提出了改进算法。首先用直方图拟合的方法选择出目标区域,然后在目标区域内进行超分辨率复原,区域外使用双线性插值。最后对3组低分辨率图像,每组五帧,用该方法进行验证。从实验结果可以看出,计算速度分别提升了15.6%、45.5%和46.5%。因此,这种方法能够有效地缩短超分辨率复原算法的处理时间。

Infrared dim-small target super-resolution restoration algorithm based on region selection

The infrared imaging technology has become the main stream research direction of the intelligent photoelectrical detection due to its advantages. However, the image of infrared dim-small target has less details and low signal to noise ratio, so super-resolution restoration algorithm is used to provide more detail information. This paper analyzes the basic principles of POCS (Projection Onto Convex Sets), for its characteristics of long running time, an improved algorithm is proposed. First, we use histogram fitting to select the target area, then use super-resolution restoration in the target area and bilinear interpolation outside the region. Finally, three groups of low resolution images with five frames in each group are used to test the performance of the proposed method. As can be seen from the experimental results, the calculation speed of the proposed algorithm increased about 15.6%, 45.5% and 46.5% respectively. In conclusion, this method can effectively reduce the processing time of the super-resolution restoration algorithm.