Stereo Camera Object Tracking via Parallax Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle safety systems using stereo cameras struggle to maintain stable and accurate tracking of objects when they are temporarily lost from view, leading to difficulties in assessing potential collisions.
Innovation Solution
An image processing apparatus and moving body device control system that utilizes a stereo camera configuration with an image analysis unit to generate parallax images, calculate distances using triangulation, and maintain object tracking through advanced image processing techniques, including parallax image interpolation and V-map generation, to ensure continuous and accurate object detection and tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a tracking technique is used to detect and track objects in parallax images, then object detection accuracy is improved, but tracking stability deteriorates when objects are temporarily lost from view
Solution Approach 1:
The system performs preliminary actions by predicting the object's future position based on its movement vector before the object actually moves out of view. This allows the tracking to continue seamlessly even when the object is temporarily lost, as the system has already prepared the next expected position and can quickly resume tracking when the object reappears.
Solution Approach 2:
The system uses feedback mechanisms by continuously monitoring the object's position and movement, adjusting the tracking parameters based on detected patterns. When an object is lost, the system uses the feedback from previous tracking data to predict and maintain the object's position, ensuring tracking stability without compromising detection accuracy.
2Measurement precision
If stereo camera measurement is used to calculate distance, then collision assessment accuracy is improved, but system complexity increases
Solution Approach 1:
The system introduces an intermediary element - the movement vector - that mediates between the stereo camera measurements and the collision assessment. By calculating the object's movement vector from sequential parallax images, the system simplifies the overall complexity while maintaining accurate distance and collision assessment capabilities.
Solution Approach 2:
The system replaces complex mechanical collision avoidance systems with image processing-based tracking and prediction. By using parallax images and movement vector analysis, the system achieves accurate collision assessment without the complexity of mechanical sensors and actuators, substituting mechanical systems with optical and computational approaches.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables stable and highly accurate continuous tracking of objects, improving vehicle safety by maintaining accurate assessment of potential collisions even when objects are temporarily lost from view.
Implementation Method 1
calculate distances using triangulation
Implementation Method 2
generate parallax images
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed is an image processing apparatus that includes a generator configured to generate distribution data indicating a frequency distribution of horizontal distance values of a range image based on the range image having pixel values according to distance of an object in a plurality of captured images; a position predictor configured to predict a predicted position of the object in the distribution data with respect to a range image of a current frame, based on the distribution data with respect to range images of a plurality of previous frames generated by the generator; and a tracking unit configured to continue tracking of the object, based on a similarity between a region of the object in a previous frame and a region in the current frame with respect to the predicted position of the object.