An Innovative Chain Coding Technique for Compression Based on the Concept of Biological Reproduction: An Agent-Based Modeling Approach

Abstract

The accelerated progress of the Internet of Things (IoT) and the nature of its heterogeneous parts cause difficulties in transferring data between different components. These rapid developments make it crucial to investigate new techniques to reduce the size of the data transmitted over the channels. In this paper, we develop an original chain code for compression that is inspired by the concept of biological reproduction. The new chain code is implemented via developing an agent-based modeling simulation consisting of rabbits, carrots, and paths for the rabbits to wander. The environment in the model represents an actual bi-level image consisting of zeros and ones. In our “biological reproduction” method, a rabbit starts wandering in the virtual world to consume carrots, while its movements are tracked and recorded by the algorithm. Each rabbit’s movement is recorded based on its previous movement. Additionally, after a rabbit consumes a certain number of carrots, it gains energy and therefore reproduces another rabbit, which continues to work on the same image, and so on. Accordingly, more rabbits simultaneously work on different parts of an image, which makes the biological reproduction method advantageous over many other classical image processing approaches. The experimental results showed that the current method could outperform well-known image compression techniques, including JBIG family algorithms, on all the images used for testing.