A new chain coding mechanism for compression stimulated by a virtual environment of a predator-prey ecosystem

In this paper, the researcher introduces a new chain coding mechanism that employs a predator-prey agent-based modeling simulation with some modifications and simplifications and uses it in compression. In the proposed method, an image is ultimately represented by a virtual world consisting of three agent types: wolves, sheep, and paths. While sheep and paths do not change their locations during the program execution, wolves search for sheep to prey on. With each step, a wolf can determine its path by choosing between seven pertinent moves depending on the encountered information. The algorithm keeps track of the wolves' initial locations and their movements and uses this as a new image representation. Additionally, the researcher introduces the 'Lengthy Advance Move,' the purpose of which is to group particular consecutive codes and further reduce the chain. This, in turn, allows the researcher to experiment with different variations of the algorithm. Finally, the researcher applies arithmetic coding on the series of movements for extra compression. The experimental results reveal that the current algorithm generates higher compression ratios than many standardized algorithms, including JBIG family algorithms. Most importantly, paired-sample t-tests reveal significant differences between the findings of the wolf-sheep predation algorithm and the other algorithms used as benchmarks for comparisons. (C) 2019 Elsevier B.V. All rights reserved.