基于特征融合的改进Camshift算法研究
Research on Improved Camshift Algorithm Based on Feature-Certificate Fusion
付振兴 1陈佩江1
作者信息
摘要
针对传统的Camshift算法仅依赖于颜色特征信息,而在面临复杂背景与尺度变化情况下跟踪效果不佳的问题,提出了一种基于特征融合的改进Camshift算法.改进算法首先将HSV颜色空间中的色度与饱和度分量联合生成模板,然后与方向梯度直方图特征自适应融合,为检测跟踪提供更多的特征信息,最后引入尺度自适应来提高算法对尺度变化时的适应能力.文章采用国际常用数据集进行测试,并与改进前算法进行比较,结果显示跟踪效果较好,并比传统算法有明显优势.
Abstract
Aiming at the problems that the traditional Camshift algorithm only depends on the color feature information,the tracking effect is not good in the face of complex background and scale changes,an improved Camshift algorithm based on feature fu-sion is proposed.The algorithm first combines the chromaticity and saturation components in the HSV color space to generate a tem-plate,and then adaptively fuses with directional gradient histogram features to provide more feature information for detection and tracking.Finally,scale adaptation is introduced to improve the adaptability of the algorithm to scale changes.The international com-monly used data sets are tested and compared with the improved algorithm,the results show that the tracking effect is better,and has obvious advantages over the traditional algorithm.
关键词
Camshift算法/特征融合/尺度自适应Key words
Camshift algorithm/feature fusion/scale adaptation引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据