This paper investigates the segmentation of multi-target image based on SVM approach combining feature extraction of kernel PCA.
实验结果表明，结合核主成份分析的特征提取，支持向量机方法是一种很有前景的多目标图像分割技术。
In chapter three, we adopt the technique of video segmentation and particle filter to develop a new algorithm of multi-target tracking.
第三章介绍利用视频分割与粒子滤波实现多目标跟踪的算法。
Synthetically utilizing features coming from infrared images and visible-light images, a novel segmentation algorithm is proposed to deal with multi-target in complicated backgrounds.
综合利用红外图像灰度特征与可见光图像灰度以及分维数方差特征，提出了一种复杂背景下多目标分割算法。
应用推荐