Preliminary study on the correlation between body weight, body measurements and captured image surface areas of growing and finishing F1 crossbred pigs

Date

5-2014

Degree

Bachelor of Science in Agriculture

Major Course

Major in Animal Science

College

College of Agriculture and Food Science (CAFS)

Adviser/Committee Chair

Elpidio M. Agbisit, Jr.

Abstract

Digital image analysis is an indirect way to measure the body weight of the animal. The objectives of the study were to determine the correlation between body weight and body measurements of F1 crossbred pigs at growing, and finishing to determine the correlation between body weight and Captured Image Surface Area (CISA) involving one-dimensional, two-dimensional, and three-dimensional images of the animal and to establish the linear regression formula that can best predict the body weight of the animal. Heartgirth had the highest correlation coefficient among other body measurements against the actual body weight with r values of 0.657 and 0.786 for growing and finishing F1 crossbred pigs, respectively. In using Captured Image Surface Area (CISA), one-dimensional and two-dimensional surface area were found to have low to moderate degree of correlation with the actual body weight of growing and finishing F1 crossbred pigs while, three-dimensional surface area, specifically front-side-top view, showed the highest significant correlation with the actual body weight of growing and finishing F1 crossbred pigs with r values of 0.956 and 0.953, respectively. In general, ?Image J®? showed promising potential for monitoring the body weight of the animal, however further evaluation is needed. Future studies should set fix angle of photography, uniform placement of markers around the body of the animal, and use of specific reference points of the marker for calibration purposes.

Language

English

Location

UPLB College of Veterinary Medicine (CVM)

Call Number

LG 993.5 2014 A3 /C38

Document Type

Thesis

This document is currently not available here.

Share

COinS