Image Enhancement from Illumination Variation to Improve the Performance of Simultaneous Localization and Mapping Technique

The Simultaneous Localization and Mapping (SLAM) is a research problem that has been explored by scholars. The SLAM contribution to autonomous robot navigation has attracted the attention of many researchers and over the decades, many SLAM techniques with outstanding achievement has been presented to resolve the SLAM problem. However, there are factors such as environmental noises that can limit the performance of SLAM techniques. These environmental noises can create misunderstandings in image analysis with an erroneous effect on the robot navigation path. In this study, attention will only focus on shadow and light intensity environmental noise and in resolving this issue, filters were presented to modify the original Monte Carlo algorithm to improve its effectiveness. MATLAB was used for the simulation experiment and evaluation is based on qualitative and quantitative methods. The experimental results acquired were compared to the original Monte Carlo algorithm without filters and the results obtained were amazing. The accomplishment of the modified Monte Carlo algorithm has led to enhancement in robot path and in real-life implementation will facilitate self-directed navigation, exploration, and route planning while it decreases accident rates and injuries for human working in an unstable environment.