Stereo Vision Object Tracking Using Disparity Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for movable platforms, such as UAVs, face inefficiencies and inaccuracies in detecting and tracking objects using conventional image data processing methods, particularly when operating in dynamic environments.
Innovation Solution
The use of stereoscopic cameras to generate disparity maps, which are processed to select and track objects within a predefined 3D volume, leveraging disparity values and sensor data for real-time object detection and tracking, even when the platform is in motion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image data processing methods are used for object detection and tracking, then the system is simpler to implement, but the detection accuracy and tracking precision are insufficient in dynamic environments
Solution Approach 1:
The patent transitions from 2D image data to 3D spatial information by generating disparity maps from stereoscopic image pairs. This dimensional transformation enables depth perception and accurate distance measurement, directly improving object detection accuracy and tracking precision in dynamic environments while maintaining manageable system complexity through established stereo vision algorithms.
Solution Approach 2:
The patent introduces disparity maps as an intermediary representation between raw stereoscopic images and object detection results. These disparity maps serve as a bridge that encodes depth information in a processed format, making it easier to extract accurate object positions and distances without requiring complex direct analysis of raw image pairs.
2Productivity
If stereoscopic cameras and disparity map processing are implemented, then object detection accuracy and real-time tracking capability are improved, but the computational complexity and processing time increase
Solution Approach 1:
The patent segments the image processing task into distinct stages: capturing stereoscopic image pairs, generating disparity maps, detecting objects in the disparity maps, and tracking detected objects. This segmentation allows each stage to be optimized independently and enables parallel processing where possible, improving real-time tracking capability while managing computational complexity through modular architecture.
Solution Approach 2:
The patent performs preliminary processing by generating disparity maps from stereoscopic images before object detection. This preliminary action transforms the image data into a format that emphasizes depth information and object boundaries, making subsequent object detection and tracking more efficient and accurate while reducing the computational burden on later processing stages.
3Measurement precision
If disparity depth data and characteristic points are used for object tracking, then tracking precision and location accuracy are enhanced, but the data processing requirements and system resource consumption increase
Solution Approach 1:
The patent extracts only the essential characteristic points from objects in the disparity maps for tracking purposes, rather than processing entire objects or all image data. This extraction approach maintains high tracking location accuracy by focusing on distinctive features while significantly reducing data processing requirements and processor energy consumption on movable platforms with limited resources.
Data Source
AI summary
A method includes obtaining a disparity map based on stereoscopic image frames captured by stereoscopic cameras borne on a movable platform, determining a plurality of continuous regions in the disparity map that each includes a plurality of elements having disparity values within a predefined range, identifying a continuous sub-region including one or more elements having a highest disparity value among the elements within each continuous region as an object, and determining a distance between the object and the movable platform using at least the highest disparity value.