Human Object Tracking via Height Map Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current object detection and tracking systems in retail environments face challenges such as background clutter, occlusion, lighting conditions, and the inability to distinguish between incoming and outgoing traffic, as well as between employees and customers, while requiring low maintenance and cost-effective solutions for precise traffic counting.
Innovation Solution
A computer-implemented system using stereo image pairs to generate height maps, detect human objects, and track their movement within a facility by integrating image capturing devices with a counting system that processes raw frames to create disparity maps, height maps, and track objects, allowing for the differentiation of incoming and outgoing traffic and separating employees from customers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If monocular video streams are used for object detection and tracking, then the system cost is reduced, but the system becomes susceptible to background clutter, lighting changes, and occlusion
Solution Approach 1:
The patent transitions from 2D intensity images to 3D range/height maps by introducing the depth dimension. This dimensional change allows the system to detect objects based on their spatial position and height, making detection immune to lighting conditions, background clutter, and occlusion that plague monocular intensity-based systems.
2Reliability
If stereo or multi-sensor systems are used to generate range or height maps, then detection robustness against lighting conditions is improved, but the device complexity increases
Solution Approach 1:
The patent replaces complex mechanical multi-sensor stereo systems with a simpler time-of-flight sensor system. The time-of-flight sensors directly measure distance by timing light propagation, eliminating the need for complex calibration and processing required by traditional stereo vision systems while maintaining robust 3D detection capabilities.
3Device complexity
If background differencing methods are used in monocular systems, then object detection is simplified, but the system becomes highly susceptible to background clutter and lighting changes
Solution Approach 1:
The patent eliminates background differencing by operating in the height map domain instead of intensity domain. Objects are detected by their height above the ground plane, which is inherently independent of background appearance, lighting conditions, and shadows that cause false detections in intensity-based differencing methods.
4Measurement precision
If adaptive template matching is used for object detection, then detection accuracy is improved, but the system becomes prone to occlusion and detection drift
Solution Approach 1:
The patent replaces template matching in the intensity domain with height-based detection in the range domain. By detecting objects based on their height maps and spatial position rather than visual appearance, the system achieves both accuracy and stability without suffering from occlusion or drift issues that plague template matching approaches.
Data Source
AI summary
A method for counting and tracking defined objects includes the step of receiving subset data with a data capturing device, wherein the subset data is associated with defined objects and includes a unique identifier, an entry time, an exit time, and location data for each defined object. The method further includes the steps of receiving subset data at a counting system, counting the defined objects, tracking the defined objects, associating a location of a defined object with a predefined area, and/or generating path data by plotting X and Y coordinates for the defined object within the predefined area at sequential time periods.