Visual Analysis of Bacterial Collective Motion

In recent years, the patterns and mechanisms of bacterial swarming have inspired the development of novel micro-robots that can mimic the behavior of intelligent biological systems, providing greater flexibility and efficiency. This has become a hot topic in the field of microbiology. The analysis of swimming bacterial swarms is currently influenced by various factors, including interactions among individual bacteria, biological characteristics, and environmental conditions. Experimental observations indicate that at high cell densities, bacteria exhibit coordinated movement on a scale of 10 to 100 micrometers. This coordination results in complex flow patterns due to the interactions and feedback between bacteria and the fluid. This paper proposes two methods for analyzing the motion of bacterial populations: the pyramidal Lucas-Kanade (LK) optical flow method and the particle image velocimetry (PIV) method. The LK optical flow method reflects the overall trajectory of the bacterial population, while the PIV method provides a vector map of the velocity of the bacterial swimming. Combining these two methods allows for multilevel analysis of bacterial motion in high-noise, low-contrast videos of swimming.