Implementation of Digital Image Processing to Determine The Body Dimensions of Nile Tilapia Fish

Abstract

The present work aimed to use digital image processing techniques for identifying body dimensions of Nile
Tilapia. The total number of exoskeleton measurements made on fish body dimension are high in aquaculture as it is used
to predict the weight, feeding schedule, breeding and market value. Current fish measuring methods can be time-consuming
and susceptible to human error. In this research, an automated and effective fish measurement system with digital image
processing is proposed. It consists of taking images of the fish, segmenting the fish body from its background and finally
using contour detection to measure a set of dimensions such as length, height and width. This analysis is applied on these
dimensions to find the weight of specific fish. Overall, the results reveal that the image processing methodology may
successfully reproduce manual measurements with a great degree of accuracy and hence can be a suitable tool for
applications in aquaculture practice. This may greatly cut down on operations.