CUDA and OpenMp Implementation of Boolean Matrix Product with Applications in Visual SLAM

In this paper, the concept of ultrametric structure is intertwined with SLAM procedure. A set pre-existing transformations has been used to create a new simultaneous localization and mapping (SLAM) algorithm. We have developed two parallel algorithms that implement time-consuming Boolean space dissimilarity matrix. The resulting matrix an important input vector quantization (VQ) step in processes. These algorithms, written Compute Unified Device Architecture (CUDA) Open Multi-Processing (OpenMP) pseudo-codes, make transformation computationally feasible on real-world-size dataset. expect our newly introduced algorithm, Fast Appearance Based Mapping (FABMAP), outperform regular FABMAP2 since spaces are more clusterable than Euclidean spaces. Another scope presented research development novel measure ultrametricity, along creation Ultrametric-PAM clustering Since current measures computational time complexity order, O(n3) lower complexity, O(n2), potential significance.