Quality metrics of simulated intensity images of coherent ladar

Abstract

Image quality metrics is an important aspect in image processing of ladar. Three quality metrics, including peak signal to noise ratio (PSNR), structural similarity (SSIM) index and speckle index (SI), are utilized to evaluate the intensity images of ladar in this paper, and the influence of quality metrics caused by different carrier to noise ratio (CNR) and different structures of images are analyzed. The results show that SSIM can accurately evaluate the image quality degraded by noise, and it is basically consistent with visual evaluation. PSNR can reflect the degree of image degraded by noise but can not reflect the decrease degree of image quality. SI can evaluate the residual degree of speckle noise which can be the criterion to weigh the uniformity of target gray values.