Graduate Studies
First Advisor
Dr. Tami Brown-Brandl
Date of this Version
Summer 7-25-2023
Document Type
Article
Citation
Takoo, G.2023. Image processing approaches to assess egg laying responses of broiler breeders (Master's Thesis).Biological Systems Engineering - Dissertations, Theses, and Student Research. Department of Biological Systems Engineering, University of Nebraska, Lincoln.
Abstract
Heat stress poses a significant challenge in broiler bird production, particularly in hot and humid conditions. Due to their rapid growth and limited heat dissipation mechanisms, broiler birds are highly vulnerable to heat stress. Mitigation strategies include evaporative cooling systems, fans, misting, and insulated housing. Addressing the laborious and error-prone manual egg counting, two studies were conducted. The first study aimed to automate egg counting for broiler breeders using computer vision. Cameras (Lorex 4K, model N841A8-Z) were set up in egg troughs and bird pens, recording 4K videos at 5 frames per second from 7 am to 2 pm. In a USDA facility, 12 pens each housing 20 hens were observed. Images were extracted at 1 frame per minute for analysis. Four techniques were compared: segmentation (90% accuracy), fixed-background subtraction (84% accuracy), YOLOv5 machine learning model (78% accuracy), and template matching (66% accuracy). The second experiment focused on egg-laying timing and ventilation effects. Over 8 weeks (June to July 2021), computer vision analyzed egg-laying times under three ventilation treatments. Images were captured every minute from 8 am to 2 pm for July, yielding around 10,800 images per pen. Continuous background subtraction achieved 94% accuracy when tested on 32,000 images from a random pen per treatment. Birds with morning-tunnel ventilation laid eggs earlier, indicating better adaptation to elevated temperatures. Notably, end nest boxes were consistently preferred for egg-laying across treatments. In summary, heat stress remains a challenge in broiler production. The first study successfully automated egg counting using computer vision techniques, with segmentation outperforming other methods. The second study revealed that enhanced morning-tunnel ventilation led to earlier egg-laying, demonstrating improved heat adaptation. Consistent nest box preference suggests a noteworthy behavior pattern. These findings collectively contribute to advancing broiler production practices and welfare under challenging climatic conditions.
Advisor: Tami Brown-Brandl
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: Agricultural and Biological Systems Engineering, Under the Supervision of Professors Tami Brown-Brandl. Lincoln, Nebraska: July, 2023
Copyright © 2023 Gautam Takoo