PTZ Camera Tracking Multiple Objects via Superobject Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing camera systems struggle to efficiently track multiple moving objects in a scene, often requiring manual intervention or switching between objects, which can lead to missed events or suboptimal viewing conditions.
Innovation Solution
A system comprising multiple overview units and a PTZ-camera that detects and locates moving objects, combines positional data to form a 'superobject', and adjusts the PTZ-camera's pan, tilt, and zoom settings to maintain a single view of multiple objects, using a minimum or user-defined zoom level for optimal imaging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single PTZ camera is used to track multiple moving objects detected by overview cameras, then the system can provide detailed views of objects, but it becomes difficult to maintain continuous tracking of multiple objects when they appear in different portions of the scene
Solution Approach 1:
The surveillance scene is divided into multiple portions, each covered by a dedicated overview camera. Each overview camera independently detects objects in its portion and generates tracking requests. This segmentation allows the system to maintain continuous monitoring of multiple objects across the entire scene by distributing detection responsibilities while a single PTZ camera provides detailed views based on aggregated requests from all segments.
2Reliability
If the PTZ camera switches between multiple objects to maintain detailed views, then tracking coverage is improved, but manual intervention or complex switching logic is required which increases system complexity and may lead to missed events
Solution Approach 1:
The system employs automatic object detection and tracking algorithms in the control unit that process positional data from multiple overview cameras and autonomously generate PTZ control commands. The system self-manages the switching between objects based on detected positions and predefined rules, eliminating the need for manual intervention and reducing operational complexity while maintaining reliable tracking coverage.
3Area of stationary object
If overview cameras have wide field of view to cover large areas, then scene coverage is improved, but the ability to capture detailed images of moving objects deteriorates
Solution Approach 1:
The system merges the wide-field-of-view coverage capabilities of multiple stationary overview cameras with the high-detail imaging capability of a single PTZ camera. The overview cameras continuously monitor large areas and detect objects, while the PTZ camera is dynamically directed to specific objects to capture detailed images. This combination allows the system to simultaneously achieve broad scene coverage and high-resolution object imaging.
Data Source
Figure 1
Figure 2~3
Figure 4A~5
AI summary
A system for tracking movable objects in a scene is disclosed. The system comprises a first overview unit arranged to capture a first overview of a first portion of a scene, and a second overview unit arranged to capture a second overview of a second portion of a scene, wherein the first portion and the second portion may be partly overlapping. The system further comprises a PTZ-camera, able to pan, tilt and zoom, so as to image selected objects in the first or second portion of the scene. The first and the second overview unit are configured to detect and locate moving objects in the scene, and to transmit positional data of said moving objects to a control unit. The control unit is configured to use said positional data as input and determine if two or more movable objects may be imaged using a single view of the PTZ-camera, and to send instructions to the PTZ-camera to assume a first zoom setting and move to a first pan position and a first tilt position, so as to image said two or more movable objects concurrently.