Bus Lane Violation Detection via Trajectory Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current photo-based traffic enforcement systems are inefficient and costly due to reliance on human reviewers, lack scalability, and fail to accurately determine if a vehicle is moving or stopped in bus or bike lanes.
Innovation Solution
A system using edge devices with cameras and deep learning models to capture and analyze video frames, detect vehicles, and determine their trajectory in both image and GPS spaces, utilizing a vehicle movement classifier to accurately assess if a vehicle is moving or stopped.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If photo-based enforcement systems rely on human reviewers to validate evidence packages, then enforcement accuracy can be maintained through human judgment, but the process becomes slow, inefficient, and costly with large amounts of human effort required
Solution Approach 1:
The system enables automated self-validation of traffic evidence packages through machine learning models that automatically detect vehicles, determine bus lane occupancy, assess movement status, and generate violation citations without human reviewer intervention, making the system self-sufficient in the validation process
Solution Approach 2:
The patent replaces the mechanical human review process with automated computer vision algorithms and deep learning models that process video frames, detect vehicles, track trajectories, and determine movement status, substituting human cognitive effort with automated computational systems
2Quantity of substance
If photo-based enforcement systems use stationary cameras to capture images, then evidence can be collected, but the systems fail to accurately determine whether a vehicle is moving or stopped and lack scalability
Solution Approach 1:
The system transitions from static image capture to dynamic video analysis, using sequences of video frames to track vehicle trajectories and determine movement status, enabling the system to distinguish between moving and stopped vehicles through temporal analysis of vehicle positions
Solution Approach 2:
The patent adds the temporal dimension by analyzing video frames across time rather than single static images, and transforms trajectory data from image space to GPS space, enabling accurate determination of vehicle movement status through multi-dimensional analysis
3Productivity
If automated lane violation detection is implemented, then enforcement efficiency and scalability improve, but the system must accurately distinguish between moving and stopped vehicles to properly assess violations
Solution Approach 1:
The system segments the violation detection process into distinct analytical stages: vehicle detection in video frames, trajectory determination in image space, coordinate transformation to GPS space, and movement classification, allowing each component to be optimized independently while maintaining overall system efficiency
Solution Approach 2:
The patent introduces trajectory data as an intermediary element that bridges vehicle detection and movement classification, using tracked vehicle positions across video frames to generate trajectory information that feeds into the movement classifier, simplifying the overall detection logic
Data Source
AI summary
Disclosed herein are methods, devices, and systems for detecting bus lane moving violations. One aspect of the disclosure concerns a method comprising capturing a video showing a vehicle located in a bus lane, inputting video frames from the video to an object detection deep learning model to detect the vehicle and bound the vehicle in a vehicle bounding polygon, determining a trajectory of the vehicle in an image space of the video frames, transforming the trajectory of the vehicle in the image space into a trajectory of the vehicle in a GPS space, inputting the trajectory of the vehicle in the GPS space to a vehicle movement classifier to yield a movement class prediction and a class confidence score, and evaluating the class confidence score against a predetermined threshold based on the movement class prediction to determine whether the vehicle was moving when located in the bus lane.