Computer Engineering and Applications ›› 2017, Vol. 53 ›› Issue (22): 176-179.DOI: 10.3778/j.issn.1002-8331.1607-0302

Previous Articles     Next Articles

Saliency based image detection and segmentation method for unmanned vehicle

ZHANG Junjie1, DING Shuyan2, LI Lunbo1, ZHAO Chunxia1   

  1. 1.College of Computer Science and Engineering, Nanjing University of Science & Technology, Nanjing 210094, China
    2.School of Electronic Engineering and Optoelectronic Technology, Nanjing University of Science & Technology, Nanjing 210094, China
  • Online:2017-11-15 Published:2017-11-29

基于视觉显著性的无人车图像检测及分割方法

张俊杰1,丁淑艳2,李伦波1,赵春霞1   

  1. 1.南京理工大学 计算机科学与工程学院,南京 210094 
    2.南京理工大学 电子工程与光电技术学院,南京 210094

Abstract: Obstacle detection is an important task in the field of environmental perception for unmanned ground vehicles. Due to the huge calculation and low segmentation precision of general obstacle detection and segmentation, a novel saliency-based obstacle detection and segmentation algorithm is proposed. Firstly, the saliency map of the scene image is generated based on the frequency tuning method. Then, the radar reflection point is mapped to the saliency map by the joint calibration of the camera and the laser radar. Finally, combined with the single camera and laser radar two kinds of sensor information, through the improved image region segmentation algorithm, obstacle detection and segmentation are achieved. In order to verify the effectiveness of the algorithm, a lot of images with various obstacles in cross-country environment are captured. The experimental results show that the algorithm is reliable.

Key words: obstacle detection, saliency detection, image segmentation, unmanned ground vehicle

摘要: 障碍物检测与分割是地面无人车辆环境感知领域中一项重要的任务。针对传统障碍物检测与分割算法的计算量大、分割精度较差等问题,提出了一种基于显著性分析的障碍物检测、分割优化算法。首先,利用基于频率调谐的方法生成场景图像的显著图;然后,通过单目摄像机与激光雷达的联合标定将雷达反射点映射到显著图上;最后,结合单目摄像机和激光雷达两种传感器信息,通过改进的图像区域分割算法,实现障碍物的检测与分割。为了验证所提出算法有效性,采集多幅包含障碍物的典型越野场景图像,对该算法进行实验与仿真验证,结果证明了该算法的有效性。

关键词: 障碍检测, 显著性分析, 图像分割, 地面无人车