SLAM系统中的自适应双阈值Canny算法

A Self-adaptive Dual-threshold Canny Algorithm in SLAM Systems

传统Canny算法在使用高斯滤波进行采样时,获得的滤波结果通常呈现为模糊图像,并且需要手动设定双阈值参数,因此缺乏对不同环境下的自适应性。为解决这个问题,本文提出一种SLAM系统中的自适应双阈值Canny算法,即采用双边滤波增强Canny算法的去噪和边缘特征保留能力,并通过自适应阈值选择算法实现在不同环境下确定最优的边缘特征提取阈值。仿真对比实验的结果表明,该算法在去噪和保留图像特征两个方面均具有更强的能力。

The traditional Canny algorithm,when using Gaussian filtering for sampling,usually produces blurred images and requires manual setting of dual threshold parameters,thus lacking adaptability to different environments.To address this issue,this paper proposes an adaptive dual threshold Canny algorithm in SLAM systems,which enhances the denoising and edge feature preservation capabilities of the Canny algorithm through bilateral filtering,and determines the optimal edge feature extraction threshold in different environments through an adaptive threshold selection algorithm.The results of simulation comparative experiments indicate that the algorithm has stronger capabilities in denoising and preserving image features.