Elevation Estimation Function for Video Analytics on Uneven Terrain
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video analytics engines face challenges in accurately determining real-world feet-to-head heights and distances between humans in images, especially when the terrain is not level, as existing methods struggle to account for varying elevations.
Innovation Solution
A computing device is used to detect human positions and estimate elevations relative to a camera device with a fixed orientation and focal length, associating these elevations with image coordinates to determine a function that estimates elevation, which is then provided to a video analytics engine for subsequent image analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If video analytics engines use standard image analysis methods, then analysis speed is maintained, but measurement precision deteriorates on uneven terrain
Solution Approach 1:
The system performs preliminary elevation estimation by detecting human positions and calculating relative elevations before conducting the main video analytics processing. This pre-processing step creates an elevation map that is then used to correct subsequent measurements, allowing accurate height and distance measurements on uneven terrain without requiring complex real-time processing during the main analysis phase.
Solution Approach 2:
The system introduces humans as intermediary objects to infer terrain elevation. By detecting human positions and using their known approximate heights, the system calculates relative elevations at different image locations, creating an elevation model that mediates between the camera view and the actual terrain geometry, enabling accurate measurements without direct terrain sensing.
2Measurement precision
If the system estimates elevation for every image location, then measurement precision improves, but computational productivity decreases
Solution Approach 1:
The system performs elevation estimation only at discrete locations where humans are detected in the image, rather than calculating elevation for every pixel or location in the scene. This partial action approach provides sufficient elevation data to construct an elevation model and correct measurements, while avoiding the excessive computational burden of complete scene analysis.
Solution Approach 2:
The image processing is segmented into discrete human detection events, where elevation estimation is performed independently at each detected human location. This segmentation allows the system to process only relevant portions of the image data, maintaining productivity while accumulating sufficient information for accurate elevation modeling across the entire scene.
3Adaptability or versatility
If the camera device has adjustable orientation and focal length, then adaptability to different scenes improves, but device complexity increases
Solution Approach 1:
The system achieves adaptability to different scenes and terrains using a camera with fixed orientation and focal length. By using humans as universal reference objects and calculating relative elevations from their detected positions, the system can handle various terrains and viewing conditions without requiring adjustable camera parameters, maintaining simplicity while achieving versatility through software-based adaptation.
Data Source
AI summary
A device, method and system for estimating elevation in images from camera devices is provided. The device detects humans at respective positions in images from a camera device, the camera device having a fixed orientation and fixed focal length. The device estimates, for the humans in the images, respective elevations of the humans, relative to the camera device, at the respective positions based at least on camera device parameters defining the fixed orientation and the fixed focal length. The device associates the respective elevations with the respective positions in the images. The device determines, using the respective elevations associated with the respective positions, a function that estimates elevation in an image from the camera device, using a respective image position coordinate as an input. The device provides the function to a video analytics engine to determine relative real-world positions in subsequent images from the camera device.