The investigation of animal behavior, most notably one of the most popular insects to humans, bees, provides data that serve to advance the design of robots and computers. In a feedback, a new computer vision system that automatically analyzes the movement of bees in a hive, accelerate research and expand knowledge about the world of insects.
The system of analysis of animal movement is part of a project called BioTracking, an effort led by robotics researchers at Georgia Institute of Technology, led by Tucker Balch, an assistant professor of computing. "We believe that the language of behavior is common to animals and robots," said Balch. "This means that, potentially, could record videos of ants for a long period of time, learn what their 'agenda' and make it work on a robot."
Social insects like ants and bees are an example of the existence of a strong and successful large-scale behavior, which arises from the interaction of many simple individuals, Balch explained. This behavior can offer ideas on how to organize a cooperative colony of robots capable of performing complex operations.
To get faster to the understanding of such behavior, Balch's team developed a computer vision system that automates the analysis of the movement, an arduous and time-consuming computing. Researchers are using the system to analyze data on the sequential movements that encode information. In bees, for example, are used to locate distant sources of food. Balch research presented at the Second International Workshop on Mathematics and Algorithms of Social Insects 16 to December 17 at Georgia Tech
The system, which has an accuracy of 81.5 percent, it can automatically analyze the movement of bees and label them based on examples provided by human experts. This level of accuracy in labeling is high enough to allow researchers to build a system to accurately determine the behavior of a bee from its sequence, Balch explained.
Some bees were marked with bright colors and recorded videos of 15 minutes. The monitoring program of the vision system translates the movements of bees marked in sets of coordinates X, Y. Then a researcher selects some data segments that are used as examples of movements for the automatic analysis.
0 comments:
Post a Comment