Real-time detection, tracking and occlusion reasoning

This invention describes a system that watches from above using cameras, like on a drone, to spot and follow people or objects. It works by analyzing how things move in the video, identifying unique motion signatures that stay the same even if the object changes speed or direction, and then uses these patterns to predict where they'll go next. The claims specifically focus on analyzing aerial imagery to achieve this.

Why it matters: Since 2018, deep learning models for object detection and tracking have become significantly more robust and efficient. This makes real-time processing of complex motion features from aerial imagery, as described, much more feasible and accurate today.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedAdvanced computer vision algorithms, software logic

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10