基于伪高动态范围的图像拼接预处理方法

来源期刊:中南大学学报(自然科学版)2019年第7期

论文作者:吴亮红 孙亮 周博文 张红强 卢明

文章页码:1619 - 1626

关键词:图像增强;伪高动态范围;加速鲁棒特征;特征匹配;图像拼接

Key words:image enhancement; pseudo high dynamic range(P-HDR); speeded up robust features; feature matching; image stitching

摘    要:在恶劣照明场景下,要提高基于特征的图像拼接过程中特征匹配准确度,图像增强是关键,为此,提出一种基于伪高动态范围(pseudo high dynamic range,P-HDR)的图像增强算法作为图像拼接的预处理方法。首先,通过单幅图像生成多重虚拟亮度图像,并进行色调映射生成伪高动态范围图像;然后,采用加速鲁棒特征(speeded up robust features,SURF)算法检测预处理后待拼接图像的特征点并执行近似最近邻搜索(fast library for approximate nearest neighbors,FLANN)算法和随机抽样一致性(random sample consensus,RANSAC)算法对其特征点分别进行粗匹配和精匹配。研究结果表明,在恶劣照明环境下,所提预处理方法可以提高特征匹配准确度。

Abstract: In the extreme illumination scenario, image enhancement is a key part to improve the feature matching accuracy in feature-based image stitching. Based on this, a pseudo high dynamic range(P-HDR) image enhancement algorithm was proposed as a preprocessing method for image Mosaic. The algorithm first generated multiple virtual luminance images from a single image, and then performed tone mapping on multiple virtual luminance images to generate pseudo high dynamic range images. Thereafter, the speeded up robust features(SURF) algorithm was used to detect the feature points of the stitched image after preprocessing, and the feature points were roughly matched and finely matched by the fast library for approximate nearest neighbors(FLANN) algorithm and the random sample consensus(RANSAC) algorithm, respectively. The results show that the proposed preprocessing method can improve the feature matching accuracy in the extreme illumination.

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号