Based on the generated image objects, this chapter addresses the multi-feature selection problem in the process of image object change detection, and conducts research on the optimal selection of multiple image object features based on the genetic particle swarm optimisation algorithm. On the basis of discussing feature selection theory and the genetic particle swarm optimisation algorithm, and combining the fitness function based on the mean and variance ratio, a multi-feature optimisation step based on the genetic particle swarm is designed. It is further applied to the change detection of image objects, and the convergence, effectiveness, and efficiency of the algorithm are analysed in detail by comparing with other optimisation methods.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Image Object Multi-feature Selection

  • Qiang Chen,
  • Yunhao Chen,
  • Mingyi Du

摘要

Based on the generated image objects, this chapter addresses the multi-feature selection problem in the process of image object change detection, and conducts research on the optimal selection of multiple image object features based on the genetic particle swarm optimisation algorithm. On the basis of discussing feature selection theory and the genetic particle swarm optimisation algorithm, and combining the fitness function based on the mean and variance ratio, a multi-feature optimisation step based on the genetic particle swarm is designed. It is further applied to the change detection of image objects, and the convergence, effectiveness, and efficiency of the algorithm are analysed in detail by comparing with other optimisation methods.