Sky Region Detection in a Single Image for Autonomous Ground Robot Navigation

• DOI: 10.5772/56884
• Publication Date: 2013-10-25T14:04:34Z
• Authors: Yehu Shen; Qicong Wang
• Keywords: Laptop

The sky region in an image provides horizontal and background information for autonomous ground robots is important vision-based robot navigation. This paper proposes a detection algorithm within single based on gradient energy function optimization. Unlike most existing methods, the proposed applicable to both colour greyscale images. Firstly, of obtained. Then, optimal segmentation threshold domain calculated according optimization preliminary estimated. Finally, post-processing method applied order refine result when no appears or objects extrude from ground. Experimental results have proven that accuracy greater than 95% our test set with 1,000 images, while processing time about 150ms resolution 640×480 modern laptop using only core.