Animal Science, Department of

 

First Advisor

Ty B. Schmidt

Date of this Version

8-2018

Document Type

Article

Comments

A THESIS Presented to the Faculty of The Graduate College at the University of Nebraska In Partial Fulfillment of Requirements For the Degree of Master of Science, Major: Animal Science, Under the Supervision of Professor Ty B. Schmidt. Lincoln, Nebraska August 2018.

Copyright (c) 2018 Jessica Michelle Lancaster

Abstract

Ensuring the health and wellbeing of pigs is of the utmost importance to the swine industry. There is a need for a real-time system that can identify changes in pig activities and activity patterns to accurately identify compromised pigs. The value of a real-time system is the capability to identify compromised pigs prior to observance of visible clinical symptoms by facility personnel. Therefore, a novel computer vision depth-enabled identification and tracking (DeIT) system was evaluated. Evaluation of 10,544 randomly selected frames indicated a 93.9% accuracy rate for identifying pigs’ identity when classified by the system as standing/walking. The accuracy of activities was 99.1% lying, 96.3% standing, 99.3% walking, 86.4% in close proximity to the feeder, and 73.6% in close proximity to the waterer. The average percentage of time spent lying (77.56±1.69%), standing (8.64±1.10%), walking (2.29±0.37%), in close proximity to the feeder (9.93 ± 1.66%), and waterer (0.95±0.28%). Furthermore, m/d walked was 943.1±105.1 m. As the trial progressed, the percentage of time spent lying and time in close proximity to the feeder increased (P≤0.001; 7.46 and 3.99%, respectively). Time standing and walking decreased (9.82 and 1.53%) from wk 1 thru wk 6. Gender had no effect (P ≥ 0.10) on percent of time spent lying, standing, walking, in close proximity to the feeder, m/d walked. Barrows spent a greater (P=0.04) percentage of time than gilts in close proximity to the waterer (1.01% vs. 0.89%, respectively). Litter had no effect (P ≥ 0.10) for time spent lying, standing, in close proximity to the feeder, and waterer. There was a difference related to the percent of time walking (P=0.05) and m/d walked (P=0.05) between litters. Results indicate two significant outcomes, 1) proposed DeIT system has the capability and sensitivity to accurately identify, maintain identification, and track the activities of nursery pigs and 2) accuracy of the DeIT system provides the potential to evaluate changes in activity for an extended period of time.

Advisor: Ty B. Schmidt

Share

COinS