基于平滑相机路径的电子稳像算法研究  

Research on Electronic Image Stabilization Algorithm Based on Smooth Camera Path

作　　者：刘少波 金晅宏[1] 钟德正 Shaobo Liu;Xuanhong Jin;Dezheng Zhong(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海

出　　处：《运筹与模糊学》2024年第3期188-199,共12页Operations Research and Fuzziology

摘　　要：在视频拍摄过程中受到环境和人为等因素的影响,拍摄出的视频存在抖动问题。为了解决视频抖动问题,提出了一种基于平滑相机路径的电子稳像方法,使相机运动路径更平滑,同时自适应地对运动补偿后的黑边进行去除,从而获得稳定的视频序列。首先,对加速稳健特征(SURF)算法进行分区域优化处理,用优化后的SURF算法进行特征点检测,并采用快速视网膜关键点(FREAK)算法对检测到的特征点进行描述;然后基于K近邻匹配(KNN)算法进行特征点对粗匹配,并采用采用随机样本一致性(RANSAC)算法和前向后向匹配算法进行特征点对的二次匹配;最后采用卡尔曼滤波算法优化估计的相机路径,获得更加平滑的相机路径,并采用双线性插值方法对原始视频序列进行运动补偿,以及自适应地去除补偿后视频序列的黑边。所提算法在两类实验场景下相较于原始相机序列的运动路径更加平滑,峰值信噪比(PSNR)值分别提升了4.39dB和4.96dB,从而说明该算法对抖动视频的稳定性和质量均有一定提升。In the process of video shooting,affected by environmental and human factors,the captured video has jitter.In order to solve the problem of video jitter,an electronic image stabilization method based on smooth camera path is proposed,which makes the camera motion path smoother,and at the same time adaptively removes the black border after motion compensation,so as to obtain a stable video sequence.First of all,the speeded up robust features(SURF)algorithm is optimized in sub-regions,the optimized SURF algorithm is used for feature point detection,and the fast retina keypoint(FREAK)algorithm is used to describe the detected feature points;then the feature point pairs are roughly matched based on the K-nearest neighbor(KNN)algorithm,the secondary matching of feature point pairs is carried out with the random sample consensus(RANSAC)algorithm and the forward and backward matching algorithm;finally,use the Kalman filter algorithm to optimize the estimated camera path to obtain a smoother camera path,and use the bilinear interpolation method to perform motion compensation on the original video sequence,and adaptively remove the black edges of the compensated video sequence.Compared with the original camera sequence,the motion path of this method is smoother in two types of experi-mental scenarios,and respectively get 4.39 dB and 4.96 dB improvement in peak signal-to-noise ratio(PSNR)value.The proposed algorithm can improve the stability and quality of shaking video.

关 键 词：电子稳像 特征点检测 特征点对匹配 平滑相机路径 运动补偿 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]