Cross-sensor Object Tracking via Grid Code Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional security camera systems in buildings or stores lack collaborative processing functions, making it difficult for security guards to maintain focus and identify abnormal events or suspicious persons across multiple screens.
Innovation Solution
A cross-sensor object-space correspondence analysis method using an edge computing architecture, where image sensing devices transmit raw data and generated grid codes to a main processing device to determine object positions on a reference plane, employing grid code look-up tables and AI for efficient object tracking without calculating traditional 3D coordinates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple screens are set up for monitoring, then the coverage area increases, but the difficulty of monitoring and identifying abnormal events increases
Solution Approach 1:
The system segments the monitoring task by dividing the space into multiple grid regions, each monitored by specific image sensing devices. The main information processing device then integrates data from multiple screens and presents unified object locations and trajectories, making it easier to monitor large areas without having to track multiple separate screens manually
Solution Approach 2:
The system merges information from multiple image sensing devices and screens into a unified object tracking system. By combining detected data from multiple sources and presenting integrated object locations and trajectories on a single reference plane, the system maintains comprehensive coverage while simplifying the monitoring operation
2Measurement precision
If traditional 3D coordinate calculation is used for object localization, then the position accuracy is high, but the computational complexity and time consumption increase
Solution Approach 1:
The system extracts only the necessary information for object localization by using grid codes that directly represent spatial positions. Instead of calculating full 3D coordinates, the system extracts and processes only the grid code corresponding to each object's projection point, significantly reducing computational complexity while maintaining positioning accuracy
Solution Approach 2:
The system changes the parameter representation from traditional 3D coordinates (x, y, z) to a simplified grid code system. By mapping object positions to discrete grid codes on a reference plane, the system achieves efficient localization with reduced computational requirements while preserving the essential spatial information needed for security monitoring
3Productivity
If grid codes are generated and processed by each information processing unit, then the processing speed increases, but the system complexity increases
Solution Approach 1:
Each information processing unit performs preliminary processing by generating local grid codes and detecting object positions in advance before transmitting data to the main information processing device. This preliminary action at the edge devices accelerates overall processing speed while the standardized grid code format keeps system complexity manageable
Data Source
AI summary
A cross-sensor object-space correspondence analysis method for detecting at least one object in a space by using cooperation of a plurality of image sensing devices, the method including: the image sensing devices sending raw data or grid code data of multiple frames of sensed images to a main information processing device to determine a corresponding projection point or a moving track of each of the at least one object on a reference plane corresponding to the space, where each of the image sensing devices has an image plane, and the raw data and each of the grid code data all correspond to a time record.