Title :
Prediction of IMF Percentage of Live Cattle by Using Ultrasound Technologies with High Accuracies
Author :
Li, ChengCheng ; Zheng, Yufeng ; Kwabena, Agyepong
Author_Institution :
Dept. of Technol. Syst., East Carolina Univ., Greenville, NC, USA
Abstract :
The purpose of this study is to produce algorithms that are able to predict the intramuscular fat (IMF) percentage of live cattle. Two algorithms based on linear regression analysis and neural network models are developed. These two algorithms extract feature information from live cattle ultrasound images and calculate the predicted IMF percentage values. The results show that these algorithms perform better than the previous studies in the same field. A brief description of the data acquisition process, the ROI extraction, the mathematics of the feature selection methods, statistical analysis on P-value and correlation, and the outputs from Matlab programs is presented.
Keywords :
agriculture; data acquisition; fats; feature extraction; mathematics computing; neural nets; production engineering computing; regression analysis; ultrasonic imaging; IMF percentage prediction; Matlab program; P-value; ROI extraction; correlation; data acquisition process; feature selection method; information feature extraction; intramuscular fat; linear regression analysis; live cattle; neural network model; statistical analysis; ultrasound image; Algorithm design and analysis; Cows; Data acquisition; Data mining; Feature extraction; Linear regression; Mathematical model; Mathematics; Neural networks; Ultrasonic imaging; Image Processing: Artificial Intelligence; Linear Regression: Neural Network; Ultrasound Image;
Conference_Titel :
Information Engineering, 2009. ICIE '09. WASE International Conference on
Conference_Location :
Taiyuan, Chanxi
Print_ISBN :
978-0-7695-3679-8
DOI :
10.1109/ICIE.2009.294
