Human Congestion and Flow Visualization System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques for analyzing and visualizing human congestion and movement flow in images from surveillance cameras primarily focus on either congestion or flow, but not both simultaneously, and lack effective visualization methods to improve human flow management and security optimization.
Innovation Solution
A system that associates image positions with map positions, detects individuals in images, converts detection data into map coordinates, and displays congestion and flow data in divided regions, using projective transformation for accurate visualization and totalization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If only congestion or flow is visualized separately, then the analysis depth for each aspect is sufficient, but the ability to grasp both congestion and flow simultaneously is insufficient
Solution Approach 1:
The patent combines congestion visualization and flow visualization into a single integrated display system. The congestion map display unit and the flow map display unit work together to show both congestion states and movement patterns simultaneously, allowing comprehensive analysis of human behavior without requiring separate analysis systems.
Solution Approach 2:
The patent adds temporal dimension to the visualization by displaying flow information as movement trajectories over time, while congestion is displayed as spatial density. This multi-dimensional approach allows simultaneous visualization of both static congestion states and dynamic flow patterns in different visual dimensions.
2Measurement precision
If image positions are directly mapped to map positions without projective transformation, then the processing is simpler, but the visualization accuracy is insufficient
Solution Approach 1:
The patent introduces projective transformation as an intermediary process between image coordinate system and map coordinate system. This transformation acts as a mediator that accurately maps positions from the camera image to the floor plan map, ensuring geometric accuracy while handling the complexity of coordinate system conversion through established mathematical methods.
3Measurement precision
If detailed detection data is collected for each person, then the analysis precision is improved, but the data processing load increases
Solution Approach 1:
The patent extracts only the essential information needed for congestion and flow analysis from the detected person data. Specifically, it extracts position coordinates and movement trajectories, discarding unnecessary detailed information about each individual. This extraction approach maintains analysis precision while reducing data processing load by focusing only on relevant features.
Solution Approach 2:
The patent segments the detection data into distinct categories: congestion-related data (position and density) and flow-related data (movement trajectory). This segmentation allows parallel processing of different data types and enables efficient visualization of both aspects without requiring comprehensive processing of all possible person attributes.
Data Source
AI summary
A setting unit associates a position in an image that is imaged by a camera with a position in a map that is obtained in advance. A detection unit detects a person in the image that is imaged by the camera. A display unit converts a position of the person detected into an associated position in the map on the basis of a setting made by the setting unit, totalizes a result of human congestion state and human flow that represents human movement on the basis of the converted position, and displays a result that is obtained.