A high-speed color-based object detection algorithm for quayside crane operator assistance system

Improvement to user interface technology for port crane operators can lead to safer and more ergonomie environments for cargo transport. An accurate and responsive container-handling guidance system can increase productivity and reduces costs. In this work, a vision-based assistive system for quayside crane operator is developed for collision warning. The system applies a new object edge detection algorithm, called Edge Approaching, to achieve faster detection rate in real-time using a stand-alone embedded system that can be easily integrated to an existing crane interface. Experiments are conducted on a scaled testbed to validate the concept. The proposed algorithms significantly increase the detection rate from as compared to the conventional Canny edge detection and Hough transform method, while maintaining a high accuracy rate of 99%.