APP下载

基于Harris角点的快速帧移动多视点图像跟踪识别

2021-08-09王佳欣窦小磊

智能计算机与应用 2021年3期

王佳欣 窦小磊

摘 要: 为了提高快速帧移动多视点图像的检测跟踪识别能力,提出基于Harris角点检测的快速帧移动多视点图像跟踪识别方法。构建快速帧移动多视点图像检测和帧稳定性控制模型,采用数字特征融合技术,实现对快速帧移动多视点图像的模糊度增强和融合处理,通过多维空间参数检测技术,实现快速帧移动多视点图像的动态参数识别,结合联合滤波特征匹配方法,采用稳像控制技术,建立快速帧移动多视点图像的灰度参数分析模型,根据Harris角点检测结果,结合角点自适应定位和分块匹配滤波,实现对快速帧移动多视点图像检测和跟踪识别。测试结果表明,采用该方法进行快速帧移动多视点图像跟踪识别的精度较高,角点定位和帧稳定性较好。

关键词: Harris角点; 快速帧移动; 多视点; 图像跟踪识别; 稳像

文章编号: 2095-2163(2021)03-0022-05 中图分类号:TP391 文献标志码:A

【Abstract】In order to improve the detection, tracking and recognition ability of fast frame moving multi-view images, a fast frame moving multi-view image tracking and recognition method based on Harris corner detection is proposed. The model of image detection and frame stability control for fast frame moving multi-view images is constructed. Digital feature fusion technology is used to enhance the ambiguity and fusion of fast frame moving multi-view images. Multi-dimensional spatial parameter detection technology is used to realize dynamic parameter identification of fast frame moving multi-view images. Combined with joint filtering feature matching method and image stabilization control technology, the gray parameter analysis model of fast frame moving multi-view images is established. According to Harris corner detection results, corner adaptive positioning and block matching filtering are combined to realize fast frame moving. The test results show that this method has high accuracy, corner location and frame stability in fast frame moving multi-view image tracking and recognition.

【Key words】 Harris corner; fast frame movement; multi-viewpoint; image tracking and recognition; image stabilization

0 引 言

随着电子稳像技术的发展,采用优化图像处理和图像角点定位技术,建立快速帧移动多视点图像处理和识别模型,提高快速帧移动多视点图像处理的检测和跟踪识别能力。在快速帧移动多视点图像处理和图像模式识别设计中,需要结合计算机视觉参数识别的方法,建立快速帧移动多视点图像处理和图像识别模型,结合视点参数识别和图像技术,实现图像的稳像处理和参数识别,提高图像成像质量,研究快速帧移动多视点图像处理方法,在图像参数分析和特征检测中具有重要意义,相关的快速帧移动多视点图像处理方法研究受到人们关注[1]。……

登录APP查看全文