Integrating 3D Flash LiDAR Data with Chaotic Optimisation for dynamic targettracking

3D flash LiDAR, defined as an advanced solid-state imaging technology that captures an entire three-dimensional scene with a single laser pulse, functioning like a laser camera instead of scanning point-by-point, has been created by Advanced Scientific Concepts (ASC) alongside the NASA Langley Research Center and various robotics and automotive tech groups to solve critical challenges in real-time depth perception, motion distortion, and autonomous hazard detection. Our research presents a new framework for dynamic target tracking that integrates 3D Flash LiDAR data with chaotic optimization algorithms. Traditional tracking methods often suffer from high computational latency, noise sensitivity, and the risk of converging to local optima when predicting non-linear trajectories. To address these limitations, the proposed approach begins by capturing high-resolution spatial data using 3D flash illumination, followed by precise point cloud filtering, voxel downsampling, and dual centroid-based registration to standardize adjacent coordinate frames. Subsequently, a chaotic optimization strategy is incorporated into the state estimation loop to thoroughly explore complex search spaces and prevent premature convergence caused by sensor occlusions or environmental disturbances. The system utilizes optimized feature subsets to dynamically update target states with high fidelity. Comprehensive experiments conducted on real-world datasets demonstrate that the proposed method significantly improves tracking accuracy, reduces computational load, and achieves a higher dynamic target detection success rate than conventional heuristic methods. Quantitative results from the testing suite indicate that our approach achieved an overall multi-target tracking accuracy of 94.2%, outperforming standard unscented Kalman filter configurations by 11.4%. It reduced average frame processing time to 115 milliseconds end-to-end, satisfying real-time navigation constraints. It maintained track continuity during 800-millisecond total visual occlusions with a trajectory recovery rate of 89.6%.