Server-Based Person Tracking Across On-Vehicle Camera Videos
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques fail to conveniently detect and analyze the movement paths of individuals from videos captured by on-vehicle cameras, limiting their application in fields like marketing and operations.
Innovation Solution
An information processing system that includes vehicles equipped with cameras and a server that communicates to detect attribute and position information of individuals, specifying and transmitting movement path data across multiple vehicle videos to improve detection convenience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a technique detects information on non-moving subjects (road attributes) from on-vehicle camera videos, then road recognition is achieved, but detection of moving subjects (person movement paths) remains impossible
Solution Approach 1:
The system transitions from static object detection to dynamic object tracking by continuously analyzing video frames over time. The server detects attribute information and position information across multiple time points, enabling tracking of moving subjects while maintaining the ability to detect static subjects like roads.
Solution Approach 2:
The system adds the time dimension to traditional spatial detection. By processing video data with temporal information and generating time-series position data, the system transforms 2D spatial detection into 3D spatio-temporal tracking, enabling movement path detection while preserving road attribute recognition.
2Measurement precision
If the system processes videos from multiple vehicles to detect person information, then movement path detection capability is improved, but system complexity increases
Solution Approach 1:
The system merges video data from multiple vehicles into a unified processing framework. The server consolidates videos, position information, and attribute data from different sources, performing integrated analysis to generate comprehensive movement paths that leverage information from all participating vehicles.
Solution Approach 2:
The server acts as an intermediary that receives, processes, and coordinates data from multiple vehicles. It manages the complexity of multi-source data integration by centralizing processing operations, matching position information across vehicles, and generating unified movement path outputs.
3Loss of information
If the system detects attribute information and position information from video data, then information extraction capability is enhanced, but processing time and computational resources increase
Solution Approach 1:
The system extracts only the essential information needed for movement path detection: attribute information (for identification) and position information (for location tracking). By focusing on these key parameters and discarding redundant video data, the system achieves comprehensive information extraction while reducing processing overhead.
Solution Approach 2:
The system performs preliminary detection of attribute information and position information from video frames before conducting full movement path analysis. This preliminary processing prepares data in advance, reducing the computational burden of subsequent tracking and path generation operations.
Data Source
AI summary
An information processing system includes vehicles and a server. Each of the vehicles generates a video obtained by imaging outside scenery in association with a time and transmit the video and position information of a host vehicle at a time when the video is generated to the server. The server detects attribute information of a person in the video from the video and detects position information of the person at the time based on the position information of the host vehicle at the time when the video is generated, when the video and the position information are received from the host vehicle. The server specifies the same person appearing in two or more videos of videos respectively received from the vehicles and transmits the attribute information of the specified person and movement path information including time-series data of the position information of the person to a client.