Computer vision-based mathematics learning enhancement for children with visual impairments

Children with visual impairments face disproportionate challenges in learning mathematics. In this paper, we explore how to apply computer vision techniques to enhance mathematics learning by fusion of multisensory information for children with visually impairments. The multisensory information includes visual and audio information. Our research focuses on the mathematics education for children with visual impairments in early ages (Kindergarten - Grade 3) by using arithmetic racks. A computer vision-based algorithm is developed to detect the numbers and positions of arithmetic rack beads and the detection results are then provided to students with visual impairments as speech guidance. Preliminary results demonstrate the effectiveness and efficiency of the proposed method for detecting the number, color, position, and moving directions of the arithmetic rack beads under different situations.