摘要
摄像平台高速靠近目标时,会使成像产生从中心到边缘呈放射状径向模糊的问题,影响对目标的探测、识别与跟踪。针对这一典型的空间变化运动模糊情形,根据实际的目标离散成像过程,推导出径向模糊图像在极坐标系中的数学模型。在分析极坐标图像纹理信息几何特征的基础上,基于非局部正则化理论提出了改进的Richardson-Lucy算法,有效解决了模糊系数存在测量误差时含噪径向模糊图像的复原问题。
When the camera platform moves to the target at high speed,the corresponding target imaging is degraded by radial blur,while the blur length increases linearly with the distance from the center to the edge.The radial blur not only affects the imaging quality,but also affects the target detection,identification and tracking.Radial-blurred image restoration is a typical space-variant motion image deblurring problem.According to the degradation analysis of radial-blurred imaging,a mathematical model of the radial-blurred image is introduced in polar coordinate.By analyzing the geometry texture information of the correlative polar coordinate images,a modified Richardson-Lucy algorithm is proposed based on non-local regularization.The proposed algorithm can restore the noisy radial-blurred images effectively even the blur parameters are contaminated by measurement errors.
出处
《武汉理工大学学报》
CAS
CSCD
北大核心
2011年第4期139-143,共5页
Journal of Wuhan University of Technology
基金
国家自然科学基金(61070009)
武汉理工大学自主创新研究基金(2010-ZY-LX-022)
