Machine Vision Based Autonomous Loading Perception for Super-huge Mining Excavator

The super-huge mining excavator has a giant structure of work devices and harsh working conditions, which results in a limited view of the operator. Therefore, how to assist the operator in determining the relative position of the bucket and the dump truck during the excavator loading operations has become a prominent issue. To solve this problem, this paper proposes a relative position perception and correction scheme based on machine vision technology. First, the Scale Invariant Feature Transform (SIFT) is used to recognize the dump truck target in the image captured by the camera. Then, the positioning algorithm based on color detection is used to identify and analyze the markers on the dump truck. Finally, through simulation tests, the proposed scheme can accurately judge the relative position of the bucket and the dump truck, and give the excavator a suitable rotation signal. The research is of great significance to the development of unmanned and intelligent excavators.