This European patent introduces a clever way for cameras and tracking systems to better handle moments when an object temporarily disappears from view—like when a person walks behind a tree or a car passes behind a building. Instead of losing track completely, the system learns from past tracking failures. It records where objects typically vanish and where they tend to reappear, building a kind of “map of blind spots.” Over time, this map helps the system make smarter guesses about where a hidden object is likely to go next. It can even adapt how long it waits before giving up on a lost object or decide where to focus its efforts to re-identify something. The result: smoother, more reliable object tracking, even when visibility isn’t perfect.
Leveraging Re-identification for Occlusion Mapping
The invention proposes using the outcome of re-identification attempts to identify and map occluded areas. When an object track is lost in a specific area and is successfully resumed through re-identification in another area, this indicates the presence of an occluded area between the “lost” area and the “resumed” area.
Building a Map of Occluded Areas
The method involves creating a map that identifies connections between areas where object tracks are lost (“start areas”) and areas where they are successfully resumed (“end areas”). This map is built by observing multiple instances of object tracks being lost and resumed.
Probabilistic Mapping
The map can further indicate the probability of an object track lost in a specific start area being resumed in a particular end area. This is based on the frequency of successful re-identification events between those areas. The map can identify multiple potential end areas for a single start area and assign probabilities to each connection.
Improving Coasting
The generated map can be used to enhance the coasting process when an object becomes untrackable. By knowing the likely end area for a lost track based on the map, the system can estimate the distance the object needs to travel while occluded. This estimated distance, and potentially the object’s speed, can then be used to dynamically adjust the coasting period.
Improving Predicted Trajectory During Coasting
The map can also provide insights into typical trajectories of objects while occluded in specific areas. By determining the angle between the start and identified end areas from the map, the system can predict the object’s position more accurately during the coasting period.