Development of a Fry Counter Software for Striped Catfish [Pangasianodon hypophthalmus (Sauvage)]

  • Philippine Journal of Agricultural and Biosystems Engineering
  • Luther John Manuel University of the Philippines Los Baños
  • Pepito Bato University of the Philippines Los Baños
  • Rossana Marie Amongo University of the Philippines Los Baños
  • Delfin Suministrado University of the Philippines Los Baños
  • Kevin Yaptenco University of the Philippines Los Baños
Keywords: fry counter, machine vision, striped catfish, pangasius, fry counter software, LabVIEW

Abstract

A counting software for striped catfish [Pangasianodon hypophthalmus (Sauvage)] fry was developed by using the Vision Development Module of LabVIEW 2010 with computer camera as a sensor. To count overlapping fry, classification algorithm was developed based on support vector machine (SVM), a pattern recognition method. Image sets with fry numbers varying from 100 to 1000 were captured by the camera and features were extracted to distinguish individual fry image.

A total of 1400 sub-images with overlapping fry were randomly selected, and 700 images were used as a training set for the classification algorithm, while the remaining images were used to verify the counting workflow. Results showed that the software is 96.30% accurate. It was verified using live fry samples and yielded an average accuracy of 91.67%. The average counting speeds for sample sizes of 100, 300, 500, and 1000 were 36.7, 37.9, 40.9 and 52.2 seconds, respectively. This indicated that the fry counter using machine vision can be an accurate and fast way to count fry up to 1000 samples and can be used by fry farmers-fisherforks to significantly improve their operation.

 

Citation:

MANUEL, L. J., BATO, P., AMONGO, R. M., SUMINISTRADO, D., & YAPTENCO, K. (2019). Development of a Fry Counter Software for Striped Catfish [Pangasianodon hypophthalmus (Sauvage)]. Philippine Journal of Agricultural and Biosystems Engineering, 15(2), 13–22. https://doi.org/10.48196/015.02.2019.02 

Published
2019-12-30