A computer-implemented method for processing objects on touch screen devices is proposed. According to the method, one or more grid sizes is determined, wherein the one or more grid sizes comprise one or more text grid sizes. The one or more text grid sizes are determined by at least one of character, word, sentence, and section. Based on the one or more grid sizes, the one or more objects on the touch screen could be divided into one or more grids. The objects on the screen could be selected discontinuously by selecting grids on the screen.