Embedded Vision Systems for Small Scale Autonomous WM ROBOT

Abstract

Due to the decrease in the farming lands as a result of increase in population and industrialization, Small Scale Wheeled Mobile Robots (SSWMR) find a lot of importance in agricultural applications like urban farming, precision farming and greenhouse farming. Path planning and navigation is one of the many problems being addressed by the research community. This requires sensing, planning and acting by the SSWMR. A number of sensors like Infra Red, Sound Navigation and Ranging (SONAR), Light Detection and Ranging (LIDAR) are applied which transmit and receive signals to understand the environment by the change in the properties of the signal. Vision sensors are considered to provide more information about the environment than the above mentioned sensors. But, this information poses the challenge of computation overhead on the processing system. In early days, it was achieved by sending the grabbed images by the camera to a remote computer system through wired or wireless communication, and then the processed result is sent back to the WMR for actuation. This makes the system bulky and complex. Autonomy can be achieved better if the processing unit is placed onboard the WMR along with the camera unit. Since the SSWMR is a small sized one, the solution can be well achieved by the use of embedded system. Size constraint is one of the many criteria based on which embedded system has gained its popularity.Embedded system has a special purpose computer developed for accomplishing specific function with limited resources like space, memory and power. newline