Identification of Lane Line Using PSO Segmentation

Abstract

In this manuscript, the process of identification of lane lines using PSO segmentation is proposed. In the image segmentation process, it is observed that the outcome of each strategy adopted needs to be optimized in terms of intensity similarity about each cluster of image pixels. Many research methodologies were experimented with to find an effective optimization for image segmentation. In this article, particle swarm optimization (PSO) optimization techniques are used for segmenting lane lines that exist on roads. Gray Level Co-occurrence Matrix (GLCM) algorithm is used to pick multiple textural features from PSO segmented images. After that, the extracted textural features were fed into a Random forest algorithm used for classification that obtains 100% of accuracy.