APS星跟踪器亚像素质心定位  

Subpixel centroiding for APS star tracker

作　　者：张晨[1] 陈朝阳[1] 张金林[1] 沈绪榜[1] 

机构地区：[1]华中科技大学图像识别与人工智能研究所,图像信息处理与智能控制教育部重点实验室,湖北武汉430074

出　　处：《光电工程》2005年第6期5-8,共4页Opto-Electronic Engineering

基　　金：航天关键技术支撑项目基金

摘　　要：APS像元结构和大小制约APS星跟踪器精度,采用散焦技术使星光成像在几个相邻像素(窗)上,以窗内灰度重心作为星点质心可以得到亚像素定位精度。APS具有多种噪声源,它们直接影响质心定位精度。根据星点模型和质心误差计算模型量化分析了窗大小、星点模型方差、噪声源对APS亚像素质心定位的影响。分析结果表明5×5窗为最优窗,星点模型方差最优值为0.62个像素,精度为0.0016个像素;固定模式噪声、暗电流对质心定位影响较大,读噪声较小,提高信噪比可以改善质心定位精度。The configuration and size of APS pixel restrain the accuracy of APS star tracker. To acquire sub-pixel position information of star point in a star image, starlight is deliberately defocused over a grid of neighbouring pixels (window). The gray center in ROI (Region of interest) window is regarded as sub-pixel precision star centroid. The APS Star Tracker has many kinds of noise resources which restrain the centroiding accuracy directly. Based on the model of star point and error calculation mode, the deviation of sub-pixel centroiding due to the size of window, the variance of spot and the noises of APS are analyzed theoretically. The analysis says that a 5 × 5window is the optimal window, the optimal variance of spot is 0.62 pixel and the corresponding accuracy is 0.0016 pixel in 5 × 5 ROI window. The main contributors to the deviation of centroiding come from dark noise and fixed pattern noise while the influence of read noise is small. The centroiding accuracy can be improved with the increase of SNR.

关 键 词：有源像素传感器 星体跟踪器 亚像素 质心定位 随机噪声 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]