Computer Vision Engineer for Multi-Object Tracking PoC

I’m looking for a Computer Vision Engineer to build a small paid proof-of-concept for tracking individually marked snails from a fixed overhead camera. Initial scope: Track 5 individually marked snails from recorded video Maintain a unique ID for each snail throughout the recording Track X/Y position and movement trajectory Detect when a snail enters predefined zones Export coordinates and detected events to CSV or JSON Produce an annotated video showing IDs and trajectories Provide the source code and basic instructions to run the PoC The main technical challenge is maintaining or correctly recovering each snail’s identity when two snails move close together, overlap, or one becomes temporarily difficult to see. This first milestone is only a feasibility PoC using recorded video. Real-time processing, website integration, mobile applications, and production deployment are not required at this stage. Please include in your proposal: 1. How you would approach the tracking and individual identification problem 2. How you would handle ID switching and temporary occlusion 3. What technologies you would use 4. Examples of relevant computer vision / object tracking work 5. Your estimated delivery time The actual tracking environment will use a natural textured background similar to the attached image. Lighting can be made brighter and more uniform, but the snails may rotate freely and the visible number markers on their shells will change orientation during movement. If the PoC is successful, there may be additional paid work involving live camera tracking, event detection, and integration with an existing platform.