Multi-Robot System Map Fusion Based on Wavelet Transform

A multi-robot system can enhance the efficiency of a single task, and can thus be employed to construct large-scale environmental maps, by enabling each robot to individually build a map and then combine the information into a single map. A map fusion approach based on wavelet transform is proposed to achieve local map fusion in order to address the issue of many local maps developed at the same site. First, wavelet decomposition is applied to the local maps to obtain sub-maps representing high-frequency and low-frequency components of the map. The wavelet transform coefficients of the maps are generated in accordance with the frequency domain features of weighted ratio and weighted gradient-based fusion rules, which are introduced to the low-frequency and high-frequency coefficients, respectively. Second, the edge detection algorithm has been improved to produce a crisper edge of the fusion map. The inverse wavelet transform is then applied to finish the map fusion. According to the simulated trials and objective evaluation, the suggested strategy works.