Fast Computing Scheme for AGV Obstacle Distance Measure and Road Recognition

Abstract:

Automatic obstacle avoidance and road detection for the Automation Guiding Vehicle (AGV) need to calculate the distance, object shape parameter. This paper presents a new obstacle distance calculating method based on monocular vision. Through scene in the two different images corresponding feature points are accurately matched, according two different video frame images disparity to compute distance between AGV and obstacle. In order to accurately find feature points, this paper uses a detection algorithm based on Harris corner, combines epipolar constraint and disparity gradient for image matching. These steps accelerate measure computing results. The basis of known structural characteristics of the road presents a road image morphology algorithm to filter road image noise, combines fast threshold algorithm to achieve a set of structured road recognition guiding system. Experimental results show that the detection method can correctly recognize the structured road of interference with certain obstacle, and achieve a visual robot guiding system.

• Periodical: Advanced Materials Research (Volumes 108-111)
• Pages: 500-506
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.108-111.500
• Online since: May 2010
• Authors: Tian Ding Chen
• Keywords: AGV, Corner Detection, Distance Measure, Road Recognition, Robot Guiding

References

[1] Blais, F., Review of 20 years of range sensor development, Journal of Electronic Imaging, 13(1): pp.231-243 (2004).

[2] Gronwall, C., Gustafsson, F., Millnert, M., Ground Target Recognition Using Rectangle Estimation, IEEE Transactions on Image Processing, 15(11): pp.3400-3408 (2006).

DOI: 10.1109/tip.2006.881965

[3] A.J. Baerveldt, A vision system for object verification and localization based on local features, International Journal of Robotics and Autonomous Systems, 34(2-3): pp.83-92 (2001).

DOI: 10.1016/s0921-8890(00)00113-5

[4] A.N. Rajagopalan, S. Chaudhuri, Uma Mudenagudi, Depth estimation and image restoration using defocused stereo pairs. IEEE Transactionson Pattern Analysis and Machine Intelligence, 26(11): pp.1521-1525 (2004).

DOI: 10.1109/tpami.2004.102

[5] D. Scaramuzza, A. Martinelliand, R. Siegwart, A Flexible Technique for Accurate Omnidirectional Camera Calibration and Structure from Motion, Proceedings of IEEE International Conference on Computer Vision Systems, pp.45-45(2006).

DOI: 10.1109/icvs.2006.3

[6] H. Kwon, Y. Yoon, J.B. Park, A.C. Kak, Person Tracking with a Mobile Robot using Two Uncalibrated Independently Moving Cameras, The Proceedings of 2005 IEEE International Conference on Robotics and Automation, Barcelona, Spain, pp.2877-2883(2005).

DOI: 10.1109/robot.2005.1570550