A Proposed Fish Counting Algorithm Using Digital Image Processing Technique

Abstract

Fish product contributes a significant amount of protein demand of human nutrition and made up of about 16% of human diet all around the world. However, Fish production is one of the factors that have been a bottleneck for development of fish farming for most developing countries such as Nigeria. One of the major and time consuming task in production is providing an accurate estimate of the fingerlings to farmers. The methods of counting fingerlings in most developing countries is done manually. These manual methods are inevitably influence by inaccuracies and exposer of the fingerlings to unnecessary stress that could lead to death. This paper proposed a fingerling counting algorithm using digital image technique. To achieved this aim, a robust segmentation algorithm, feature extraction algorithm and machine learning algorithm for fingerlings classification and counting are hereby formulated. At the end of this research, the proposed algorithm is expected to count different sizes of fingerlings with high accuracy.