An automated length measurement system for tilapia fish based on image processing technique

Abstract

In aquaculture industry, measurement of fish length plays an important role in monitoring and determining fish growth, gender, age and reproduction. Currently, fish rearing industries are using measuring wooden board or acrylic plastic with the scale attached to the board itself. However, this method is causing a stress on the fish when measuring their length alive. Furthermore, manual measurement of fish length one by one using measurement board is time-consuming. Hence, this research presents an automated system that is able to measure fish length specifically tilapia fish by using image processing technique and Graphical User Interface (GUI). This automated GUI system is developed for end user and it is able to measure the length of tilapia automatically and accurately. The system is designed by using image processing tool, which consists of image segmentation and feature extraction to get the fish-shaped region. Based on the results, the GUI system is able to measure automatically and accurately the fish length based on the fish image. This developed system resulting with a 8.50% mean relative error and 91.50% accuracy of the tilapia fish length measurement for a 20 cm distance between the digital camera and the tilapia fish. These results show that the proposed system is capable to produce faster results and yield accurate length measurement of the tilapia fish. Thus, with this automated system, it is potentially assisting in fish reproduction and growth for aquaculture industry especially for tilapia industry.