DASA : an open-source design, analysis and simulation framework for automotive image-based control systems

Image-Based Control (IBC) systems are a class of data-intensive feedback control systems whose feedback is provided by image-based sensing using a camera. IBC has become popular with the advent of efficient image processing systems and low-cost CMOS cameras with high resolution. The combination of the camera and image processing (sensing) gives necessary information on parameters such as relative position, geometry, relative distance, depth perception and tracking of the object-of-interest. This enables the effective use of low-cost camera sensors to enable new functionality or replace expensive sensors in cost-sensitive industries like automotive. The state-of-the-art design, analysis, and simulation of IBC assumes that the sensing algorithm is executing correctly with an assumed or estimated worst-case delay. The sensing algorithm is simulated and validated using static pre-captured image streams and is normally decoupled from the control algorithm. However, in reality, the camera is fixed to the vehicle body and any steering change would affect the region captured by the image. This dynamism cannot be captured in a static image stream and a dynamic image stream that considers the change in vehicle dynamics due to IBC actuation is needed. We present an open-source design, analysis, and simulation framework for automotive IBC systems that can consider the change in vehicle dynamics in real-time and produces real-time dynamic image stream as per the control algorithm. Our framework models the 3D environment in 3ds Max, simulates the vehicle dynamics, camera position, environment and traffic in V-REP and computes the control output in Matlab. Our framework runs Matlab as a server and V-REP as a client in synchronous mode. We show the effectiveness of our framework using a vision-based lateral control system.