基于梯度图的快速POCS超分辨率复原算法研究

随着红外成像相关产业的兴起,红外成像技术具有的隐蔽性好、探测范围广、定位精度高、穿透距离远,以及轻质小巧、低耗可靠等优点备受青睐,已成为当前智能化光电探测发展的主流方向。然而,红外弱小目标的图像细节特征少、信噪比低等特点成为红外图像应用的瓶颈,如何提高红外弱小目标成像效果成为目前的研究热点。POCS算法是目前超分辨率复原中应用非常广泛的一种复原算法,但是该算法运算量大,处理时间较长,同时对图像的边缘细节保留能力较差。针对POCS超分辨率复原算法迭代时间较长,无法满足光电探测系统实时性的问题,提出了基于梯度图的快速POCS超分辨率复原算法(GPOCS)。该算法根据图像的梯度分布对图像中的像素点进行分类,采用不同的迭代系数进行计算。改进算法能够较好的保留边缘信息并抑制噪声,进而在保证超分辨率复原性能的基础上大大缩短了运算时间。实验结果表明:GPOCS算法复原结果在背景处噪声得到一定的抑制,整体复原能力优于传统的POCS复原方法。该算法能够有效地保留边缘细节,同时处理时间小于传统的POCS复原方法,减少了1个数量级已经是接近实时。GPOCS算法能够自适应的选取迭代步长,较好的保留边缘信息并抑制噪声,进而在保证超分辨率复原性能的基础上大大缩短了运算时间,虽然不能满足实时性的要求,但是也已经是接近实时。

Research on fast POCS super-resolution restoration algortihm based on gradient image

With the spring up of the infrared imaging related industry, the infrared imaging technology has become the mainstream development direction of the intelligent photoelectrical detection due to its good concealment, wide detection range, high positioning accuracy, long penetration distance, light weight, little volume, low power dissipation and high solidity. However, the features of the image of infrared dim-small target, such as less details and low SNR, become the bottleneck of the application of infrared image. How to enhance the imaging effect of the infrared dim-small target becomes the hotspot of the research. POCS algorithm is currently one of the widely used super resolution restoration algorithm. However, this algorithm requires large amount of computation and takes a long processing time. Also, the retention capacity of the details on the edge of the image is poor. Aiming at the long iteration time of the POCS super-resolution restoration algorithm that cannot meet the real time detecting requirement of optical detection system, a fast POCS super resolution restoration algorithm based on gradient image (GPOCS) is proposed, which classifies the image pixels according to the gradient distribution of the image, and then uses different iteration factors for calculation. The iteration step is larger when the gradient is larger and the iteration step is smaller when the gradient is smaller. The improved algorithm can preserve edge information and suppress noise. Therefore, it can guarantee the performance of the super-resolution restoration and greatly reduce the operation time. Experiment results show that GPOCS algorithm results in certain noise suppression at background. Its overall restoration capability is superior to that of traditional POCS method. This algorithm could effectively retain the edge details, and the processing time is less than that of traditional POCS restoration method; and the one order of magnitude reduction is already close to real time performance. The GPOCS algorithm could adaptively select the step size. GPOCS algorithm could better retain edge information and suppress noise. Furthermore, the GPOCS algorithm could guarantee the super-resolution restoration performance, while greatly reducing the processing time. Although GPOCS algorithm could not meet real time requirement, its performance is already close to real time.