AI-Based Sheep Counting Software

# AI-Based Sheep Counting Software (YOLO / Computer Vision) ## Project Overview I am looking for an experienced Computer Vision / AI developer to create software for automatically counting sheep entering and leaving a livestock enclosure. **Important:** I do NOT need hardware installation, camera setup, networking, or computer configuration. I will handle all hardware procurement, camera installation, and on-site setup myself. I only require the software development. This is a real-world livestock monitoring project and reliability is extremely important. Accuracy is more important than speed. I prefer a robust and reliable counting system over a highly complex solution. ## Project Description I own a flock of approximately 1,000 sheep. Every day, the flock leaves through the same corridor in the morning and returns through the same corridor in the evening. The sheep are all white and are not individually identifiable. I do NOT need individual animal identification or tracking of specific animals. The only objective is to determine whether the same number of sheep returned as left. ### Example Morning Exit Count: 1,003 Evening Return Count: 998 Missing Sheep: 5 If the numbers do not match, the software should automatically generate an alert. This is critical because any sheep left outside overnight may be attacked by predators, so immediate notification is required. ## Technical Requirements * Python preferred * YOLO-based object detection * Object tracking (ByteTrack, DeepSORT, or similar) * Virtual line-crossing counter * Support for one or multiple IP cameras * Real-time operation * Local processing (no cloud dependency) * Telegram notification support (preferred) * Daily count logging * Ability to operate continuously during the grazing season ## Expected Features * Count sheep leaving the enclosure * Count sheep returning to the enclosure * Compare exit and return counts * Generate automatic alerts when counts differ * Store historical daily records * Provide a simple configuration interface * Adjustable confidence thresholds * Adjustable counting zones and virtual lines ## What I Will Provide * Sample videos from the actual farm * Camera specifications * Corridor dimensions and layout * Real-world test footage * Computing hardware * On-site testing and feedback ## Deliverables * Complete source code * Installation documentation * Configuration guide * Deployment instructions * Ability to run locally on my hardware ## Preferred Experience Please apply only if you have experience with one or more of the following: * YOLO * OpenCV * ByteTrack * DeepSORT * Object counting systems * Real-time video analytics * Livestock monitoring * Smart agriculture solutions ## When Applying Please include: 1. Similar projects you have completed 2. Relevant portfolio examples 3. Estimated development timeline 4. Fixed-price quotation 5. Recommended software architecture 6. Expected counting accuracy under real-world conditions This project may expand into additional livestock monitoring features in the future, so I am interested in establishing a long-term relationship with the right developer.