Abstract: Chunks of coal gangue in the course of carriage jam serious easy cause equipment wear and tear,in order to realize the coal transfer in the process of waste rock fragmentation jam,large machine caused by the problem of large equipment maintenance time is long,input,using nonlinear photographic imaging model and K means clustering algorithm for image acquisition and intelligent robot obstacle recognition was studied,and realized with coal waste more body shape,regional characteristics of nonlinear camera and accurate extraction,provides an important reference for the application of intelligent robot,to solve the problems in the process of coal transfer has practical significance.

Key words: intelligent robot,nonlinear photographic imaging,Kmeans clustering algorithm