Several technical challenges occur during palmprint capture and prepro cessing on mobile devices,including non-uniform palm location and gesture,complex background,d iverse illumination,limited hardware resource,etc.Most methods have low accuracy an d are prone to failure in the above scene.Double-point auxiliary acquirement technique is developed to overcome the aforementioned problems,in w hich two auxiliary points locate on two finger valleys.The centroids of the two square restriction blocks are the two auxiliary points.Palm region in each restriction block is segmented based o n the adaptive Gaussian skin-color model that is trained in the sub-region,i.e.,the rectang le at the bottom of the restriction block.The uncompleted palm region is segmented by the likeli hood threshold. The adaptive Gaussian skin-color model is retrained in the uncompleted palm re gion,and finally the palm region is accurately segmented by the newly set likelihood threshold an d the effect is better than the first segmentation. The developed method reduces the computation complexity of palmprint preprocessing,and accordingly improves real-time performance.The expe rimental results confirm that the developed method not only remarkably reduces the failure rate of palm segmentation in comp lex background,but also achieves higher verification accuracy in mobile environm ents.