Jul 08th, 2016


This summer, rising sophomore Spencer Ader is working with Dr. DeWitt, Ty Carlson, and Max Maurer on developing a self-driving RC car as part of HPU’s Summer Undergraduate Research Program in the Sciences (SuRPS). His primary contribution to the project deals with the system’s computer vision — he is using back-projected histograms to isolate objects of interest in the field of view. Their first major task in computer vision was to follow a yellow ball, as seen in the left panel of the second image attached. The colored histogram of the yellow ball takes the pixel color in HSV (Hue, Saturation, and Value) and plots the number of pixels as a function of each HSV combination. This is important to the visual analysis so they can filter out unwanted colors and isolate the object of interest. This is achieved by comparing each pixel in HSV of the picture on the left to a histogram of the yellow ball. If the HSV values match then a black dot is represented; if not, a white dot is represented. This is shown in the right panel of the second image. When the ball is detected, the steering is corrected to keep the ball in the center of the screen. They have had great success so far, so we’re looking forward to seeing what all the RC car can do by the end of SuRPS!