An automatic identification method for the cane-end of sugarcane based on digital image processing is given considering the characteristic of the cane-end of sugarcane. An image processing method with Visual C++ is used in the dynamic acquisition and jitter procedure, and identifying the cane-end automatically. The identification system of the cane-end of sugarcane consists of the image acquisition unit, the image processing unit, the identification and control unit. After the digital image of a cane-end is collected, the identification system locates the edge between the cane-end and the body of the sugarcane immediately and automatically via image processing and identification, providing the location information of the cane-end for the following mechanical cane-end cutting method. This system can massively improve the productivity of the cane-end cutting mechanism in a sugarcane harvester.