A5086100243- Image Processing Techniques and Applications
- Effects of Environmental Stressors on Livestock
- Smart Agriculture and AI
- Advanced Measurement and Detection Methods
- Industrial Vision Systems and Defect Detection
- Animal Nutrition and Physiology
- Animal Behavior and Welfare Studies
- Advanced Vision and Imaging
- Human Pose and Action Recognition
- Hand Gesture Recognition Systems
- 3D Shape Modeling and Analysis
- Genetic and phenotypic traits in livestock
- Meat and Animal Product Quality
- Advanced Chemical Sensor Technologies
- Remote Sensing and Land Use
- Advanced Image Processing Techniques
- Animal Virus Infections Studies
- Advanced Image and Video Retrieval Techniques
- Textile materials and evaluations
- Robotic Locomotion and Control
South China Agricultural University
2022-2025
Poultry managers can better understand the state of poultry through behavior analysis. As one key steps in analysis, accurate estimation posture is focus this research. This study mainly analyzes a top-down pose method multiple chickens. Therefore, we propose “multi-chicken pose” (MCP), system for chickens deep learning. Firstly, find position each chicken from image via detector; then, an estimate made using network, which based on transfer On basis, pixel error (PE), root mean square...
At present, raising caged chickens is a common farming method in China. However, monitoring the status of still done by human labor, which time-consuming and laborious. This paper proposed posture detection for based on computer vision, can automatically identify standing lying cage. For this aim, an image correction was used to rotate make feeding trough horizontal image. The variance speeded-up robust features were indirectly obtain key area through position. In paper, depth camera...
Accurately counting chickens in densely packed cages is a major challenge large-scale poultry farms. Traditional manual methods are labor-intensive, costly, and prone to errors due worker fatigue. Furthermore, current deep learning models often struggle with accuracy caged environments because they not well-equipped handle occlusions. In response, we propose the You Only Look Once-Chicken Counting Algorithm (YOLO-CCA). YOLO-CCA improves YOLOv8-small model by integrating CoordAttention...
Poultry pose estimation is a prerequisite for evaluating abnormal behavior and disease prediction in poultry. Accurate pose-estimation enables poultry producers to better manage their Because chickens are group-fed, how achieve automatic recognition has become problematic point accurate monitoring large-scale farms. To this end, based on computer vision technology, paper uses deep neural network (DNN) technique estimate the posture of single broiler chicken. This method compared detection...
Abstract Accurate poultry detection is crucial for studying behavior using computer vision and video surveillance. However, in free-range farming environments, detecting chickens can often be challenging due to their small size mutual occlusion. The current algorithms exhibit a low level of accuracy, with high probability false missed detections. To address this, we proposed multi-object chicken method named Super-resolution Chicken Detection, which utilizes super-resolution fusion...
The speed and accuracy of navigation road extraction driving stability affect the inspection cage chicken coop robots. In this paper, a new grayscale factor (4B-3R-2G) was proposed to achieve fast accurate extraction, line fitting algorithm based on boundary features improve algorithm. achieved 92.918% segmentation accuracy, six times faster than deep learning model. experimental results showed that at 0.348 m/s, maximum deviation visual 4 cm, average 1.561 acceleration 1.122 m/s