A Novel Lanemark Identification System using Kalman Filtration Logic and Image Processing Principles

Lane-mark extraction utilizing visual cognitive computing is one of most important components of advanced driver assistance systems in intelligent transportation systems that are used for autonomous driving. In order to extract lane markings from road scenes, driver less cars use onboard cameras positioned on the front. There are four primary aspects to our novel approach for extracting lane markers. On begin; this article applies a gray-scale and quick median filter to road photos acquired by onboard cameras. Using the lane mark characteristics as constraints, we then present a multiconstraint lane-features filter for extracting lane markings. Clustering features is accomplished via the use of a p-least squares algorithm based on the double point removal method, and prospective lanes are discovered through the use of a recursive dichotomy algorithm. The next step is to validate and refine probable lane markers so that we may achieve more accurate and dependable data extraction findings. Our research focused on four categories of often complex traffic scenarios. Experiments have shown that the suggested technique can reliably extract lane markers from complicated real-world environments. In this study, a lane-mark extraction evaluation approach is proposed as well as partial test results have been reviewed.