Crowd Interest Detection Using Depth Image Height-Top-View Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting the interest degree of a crowd in a target position are subjective and inaccurate, relying solely on crowd density without considering motion and orientation, and are challenging in crowded areas where individual tracking is difficult.
Innovation Solution
A method that projects a depth image onto a height-top-view, divides it into cells, determines crowd density, moving speed, and orientation, and calculates interest degree based on these factors using a density model and statistical learning, enabling objective and accurate detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual count is used to determine crowd density, then accuracy is improved, but human cost increases
Solution Approach 1:
The patent replaces manual mechanical counting with an automatic image processing system that uses depth images and statistical learning to detect and count crowd density, eliminating human labor while maintaining high accuracy through automated analysis of height-top-views and density maps
Solution Approach 2:
The system performs self-service by automatically capturing depth images, processing them through the density detection model, and generating crowd density measurements without requiring human operators, enabling continuous autonomous monitoring of crowd conditions
2Loss of time
If automatic count based on WIFI or RFID is used, then human cost is reduced, but accuracy decreases
Solution Approach 1:
The patent transitions from two-dimensional RFID/WIFI signal-based counting to three-dimensional depth image analysis with height-top-view projection, adding vertical dimension information that enables more accurate person detection and counting while maintaining automatic operation
Solution Approach 2:
The system changes the detection parameter from signal presence/absence in RFID/WIFI to physical height and depth information in images, using the height-top-view projection and density detection features that leverage vertical dimension parameters to improve counting accuracy
3Measurement precision
If individual tracking in photographed image is used, then accuracy is improved, but accuracy greatly decreases at crowded places
Solution Approach 1:
The patent extracts only the necessary height and depth features from individual persons to create density detection features, rather than attempting to track complete individual trajectories, thereby maintaining detection capability in crowded environments where full tracking becomes infeasible
Solution Approach 2:
The system performs partial action by detecting and counting persons based on height-top-view projections and density features without completing full individual tracking, which is sufficient for crowd density measurement and remains effective even when complete tracking would fail in dense crowds
4Device complexity
If only crowd density is considered to determine interest degree, then simplicity is improved, but objectivity decreases
Solution Approach 1:
The patent merges multiple factors including crowd density, motion characteristics, and orientation information to comprehensively determine interest degree, combining these elements through weighted integration to achieve objective and accurate assessment while maintaining systematic simplicity
Data Source
AI summary
A method and an apparatus for detecting an interest degree of a crowd in a target position are disclosed. The interest degree detection method includes projecting a depth image obtained by photographing onto a height-top-view, the depth image including the crowd and the target position; dividing the height-top-view into cells; determining density of the crowd in each cell; determining a moving speed and a moving direction of the crowd in each cell; determining orientation of the crowd in each cell; and determining, based on the density, the moving speed, the moving direction and the orientation of the crowd, the interest degree of the crowd in each cell in the target position. According to this method, the interest degree of the crowd in the target position can be detected accurately, even at a crowded place where it is difficult to detect and track a single person.