Modeling Ecosystem Management Based on the Integration of Image Analysis and Human Subjective Evaluation - Case Studies with Synecological Farming

Abstract

The challenge for biodiversity restoration and augmentation is to find effective indicators for ecosystem management without discarding too much of the complexity that contributes to functionality. Many technical challenges lie ahead in setting up information measures to manage dynamically changing ecosystems in the real world. It is expected that image analysis features such as edge, texture, color distribution, etc. will provide clues, but methods to evaluate their effectiveness in the context of integrated management have not been sufficiently studied. Taking synecological farming (Synecoculture™) as a typical example of complex ecosystem management, we investigate the initial steps toward the construction of an evaluation model by incorporating image analysis and empirical knowledge acquired by human managers. As a result, we showed that it is possible to construct a model that connects the features of image analysis and human subjective evaluation with consistency according to the level of the evaluators and proposed a cycle that would refine both the evaluation model and associated human capacity. We also presented an interface for utilizing collective knowledge in ecosystem management using the proposed model and the prospect of scaling up in conjunction with robotics.