Open AccessPrediction of cow’s body weight by measurement of the body dimensions with image analysis
Author(s) -
I Wayan Astika,
Rahmat Hidayat
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/542/1/012070
Subject(s) - circumference , mathematics , linear regression , regression analysis , statistics , body height , stepwise regression , body weight , geometry , medicine
Measurement of cow’s body dimensions manually for the prediction of cow’s body weight are less practical. The purpose of this research is to determine the body length, chest circumference, hump height, and hip height for prediction of the mass of cow’s body with image processing. Prediction of the body dimensions with image processing was done by drawing a line on the image correspond with the point of body dimensions measurement and then calculating the number of pixels of each variable of the body size with the Euclidean equation. The mass of cow’s body is predicted by using multiple linear regression equations with two models of regression equation i.e. 1) regression equation with two variables (body length and chest circumference) and 2) regression equation with four variables (body length, chest circumference, hump height, and hip height). Dimensions of cow’s body obtained from image processing were usually smaller than those obtained from manual measurement. As compared to manual measurements, the average error of body length was 8.8%, the average error of chest circumference was 4.0%, the average error of hump height was 11.8%, and the average error of hip height was 8.3%. Based on regression models, hump height and hip height did not much affect the cow’s body weight, so the regression models only use with two variables: body length (BL) and chest circumference (CC). The model gave an average error 6.4% or about 17.5 kg out of the average weight 234 kg.