Stereoscopic fish length measurements using neural networks

Authors

  • Carlo M. Orofeo Department of Physics, University of San Carlos
  • Rhodesa U. Cruzet Department of Physics, University of San Carlos
  • Michael Paul B. Kindica Department of Physics, University of San Carlos
  • Ben Zuidberg Department of Physics, University of San Carlos
  • Kees Karremans Vrije Universiteit Amsterdam, The Netherlands

Abstract

An existing stereoscopic technique employing neural networks has been used to measure the length of fishes.
Investigations prior to the actual measurements show that external parameters such as salinity and the orientation of the object with respect to the cameras have negligible effect on the measurements. With a total error less than 2 mm, our method is far better than conventional techniques.

Downloads

Issue

Article ID

SPP-2002-1F-04

Section

Image and Signal Processing

Published

2002-10-23

How to Cite

[1]
CM Orofeo, RU Cruzet, MPB Kindica, B Zuidberg, and K Karremans, Stereoscopic fish length measurements using neural networks, Proceedings of the Samahang Pisika ng Pilipinas 20, SPP-2002-1F-04 (2002). URL: https://proceedings.spp-online.org/article/view/SPP-2002-1F-04.