Adltformer Team-Training with Detr: Enhancing Cattle Detection in Non-Ideal Lighting Conditions Through Adaptive Image Enhancement
Zhiqiang Zheng, Mengbo Wang, Xiaoyu Zhao, Zhi Weng · 2024
Read Adltformer Team-Training with Detr: Enhancing Cattle Detection in Non-Ideal Lighting Conditions Through Adaptive Image Enhancement by Zhiqiang Zheng, Mengbo Wang, Xiaoyu Zhao, Zhi Weng (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11672519.txt.
Description
Zhiqiang Zheng, Mengbo Wang, Xiaoyu Zhao, and Zhi Weng published a study in 2024 on a system for detecting cattle in farm images taken under poor lighting conditions. Backlighting, uneven lighting, and low light can wash out detail, distort color, and add noise. That degrades the visual quality of images and the accuracy of detection models.
The team's approach pairs an image enhancement model with a detection model and trains them together as a team. To teach the enhancer, they synthesized low-light images from daylight images, creating aligned pairs with controlled noise levels. The enhanced images feed the detector in a form suited to machine vision. In experiments, the paired system outperformed the detector on its own and a sequential enhance-then-detect pipeline. It achieved a detection score under difficult lighting while keeping its speed and size in line. The work is aimed at farm use.
Catalog record
| CREATOR | Zhiqiang Zheng, Mengbo Wang, Xiaoyu Zhao, Zhi Weng |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |