The palm is one of the biometric characteristics that has been relatively recently investigated for identification and verification systems. The reason for using the palm geometry feature is, because the palm geometry is considered more resistant to external factors, such as weather, dry or wet palm conditions compared to using the characteristics of the palm lines that have difficult details and are susceptible to external factors. The problem that often arises in the self-recognition system is that it is easy to commit a crime against a person's identity if only by using something that is owned or something that is known to a system, using biometrics techniques is expected to minimize these frequent problems. Therefore, this study was made to implement the region of interest (ROI) segmentation method for palm line imagery using sobel edge detection, so that it can help for the initial process of identification and verification. the highest accuracy value on the right palm line image reached 87.01% and the lowest reached 86.46%, the highest accuracy value on the left palm line image reached 85.35% and the lowest reached 82.68%.