Classifying Mangrove Crub Images for Growth Stages Detection and Monitoring

Abstract

This is a research-in-progress of designing an intelligent system for mangrove crab larval growth stages development characterization and detection. This research applies image processing, machine learning, and prototyping in the design of the system. An initial experiment is conducted to verify the accuracy of classification and recognition. The model achieved an average of 85% accuracy in classification of larval images samples. This study contributes to the development of the corpus of mangrove crab larval images in a context of a developing country. This paper also recommends further enhancement of the system.