The design of DSP-based corner detection for power line inspection robot

Aiming at the requirement of power line inspection robot, this paper develops and builds a vision system depending upon the TMS320DM642 as the core, which can realize the image collection and display, implement the image features extraction. Through the extended UART interfaces, this system can communicate with the control unit of Power Line Inspection Robot. Corner coordinates obtained from the algorithm of this system can be used for extraction and matching of the image feature points, motion estimation, three-dimensional reconstruction, etc. Compared to the traditional vision system based on image capturing cards and PC platforms, it has the advantages of small size, favourable real-time properties, low power consumption, etc. Experiment results indicate that the precision and speed of corner extraction are both better than those of the traditional PC platforms.

[1]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[2]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[3]  Christopher G. Harris,et al.  A Combined Corner and Edge Detector , 1988, Alvey Vision Conference.