The proportion of internal foreign matter is the largest in the insulation fault of GIS equipment

and whenthe foreign objects inside the long-tube structure GIS equipment are detected

the traditional vacuum cleaner clean-ing and manual wiping methods are difficult to play an effective role. Based on the intelligent inspection and cleaningrobot platform of the GIS equipment

this paper designs the dual-light source light-filling technology which relies onthe combination of ultraviolet light source and visible light source in order to clean the foreign objects in the GIS pipe-line more effectively

and studies the foreign matter recognition algorithm to process images taken by robot using dual-light-source compensation

which can be used as a completely new scheme to check and verify the foreign matter de-fects of GIS equipment and to clean up foreign matter. Under the two kinds of light sources

the images taken insidethe pipeline are processed to make foreign matter recognition judgment

which provides guidance for the robot toclean foreign matter. In this paper

while realizing the visualization of the foreign matter inside the GIS cavity

theproblem that the single light source of the intelligent inspection and cleaning robot inside the GIS equipment is diffi-cult to find some foreign particles is solved. It also avoids the problem that the robot needs manual experience to dis-criminate the foreign object

and improves the efficiency of the foreign object inspection inside the GIS equipment.