Vehicle Yaw Orientation Representation Using Normalized Split-Line Shift
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for representing the yaw orientation of vehicles in images are either too complex or unstable, leading to tracking inaccuracies due to the use of binary values for split lines that can jump between sides, especially when vehicles are viewed from different angles.
Innovation Solution
A method and device for representing the yaw orientation using a signed horizontal shift of a vertical split line within a bounding box, normalized by the bounding box width, to stabilize vehicle tracking across image frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If binary values are used for split lines to indicate vehicle portions, then the representation is simple, but tracking stability deteriorates due to jumps between sides when vehicles are viewed from different angles
Solution Approach 1:
The patent changes the parameter representation from discrete binary values to continuous normalized horizontal shift values. Instead of using binary indicators that can jump between states, the invention uses a continuous parameter (normalized shift θ) that varies smoothly with vehicle orientation, eliminating abrupt transitions and improving tracking stability.
Solution Approach 2:
The patent transitions from a one-dimensional binary representation to a two-dimensional normalized coordinate system. By introducing the horizontal shift as a normalized value between -1 and 1, the system adds dimensional information that captures vehicle orientation more accurately without proportionally increasing complexity.
2Reliability
If normalized horizontal shift is used to represent yaw orientation, then tracking stability improves, but computational complexity increases due to normalization calculations
Solution Approach 1:
The patent transforms the raw horizontal shift parameter into a normalized parameter θ by dividing by the bounding box width. This parameter transformation improves tracking stability by creating a scale-invariant representation that works consistently across vehicles of different sizes while maintaining computational efficiency through a simple division operation.
3Measurement precision
If 3D bounding boxes with six degrees of freedom are used for vehicle representation, then orientation accuracy improves, but computing load increases due to multiple parameters
Solution Approach 1:
The patent extracts only the essential yaw orientation information from the full 6DOF 3D bounding box representation. By focusing specifically on the horizontal split line position and its normalized shift, the invention isolates the critical orientation parameter needed for tracking, discarding redundant parameters and reducing computational load while maintaining orientation accuracy.
Solution Approach 2:
The patent projects the three-dimensional vehicle orientation problem onto a two-dimensional image plane, where yaw orientation can be represented by the horizontal position of the vertical split line. This dimensionality reduction simplifies the representation from six parameters to a single normalized horizontal shift value, significantly reducing computing load.
Data Source
AI summary
A device and a method for representing the yaw orientation of a vehicle visible on an image, including: obtaining the position in the image of a rectangular bounding box surrounding the vehicle, obtaining the position of a vertical split line of the bounding box configured to be placed at a corner of the vehicle separating a first portion of the vehicle from a second portion of the vehicle, the first and second portion being either the front, the back, or a side of the vehicle, determining a signed horizontal shift of a horizontal position, associated with the vertical split line, with respect to a horizontal reference position, associated to the bounding box, normalizing the signed horizontal shift to obtain a value θ to be used in the representation of the pose of the vehicle.