Optical Flow Processing for Low-Cost Cross Traffic Alert
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Cross Traffic Alert (CTA) systems are too expensive for the automotive mass market, requiring external image processors and in-vehicle sensors for heavy computation, and previous low-cost solutions using single image cameras are not effective in detecting obstructing vehicles without overloading the Engine Control Unit (ECU).
Innovation Solution
A low-cost CTA method utilizing Optical Flow (OF) data processed directly within the Image Signal Processor (ISP) of a camera, filtering Motion Vectors to calculate the Vanishing Point and apply a Horizontal Filter, reducing computational load and eliminating the need for external processors or additional sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If external image processors and in-vehicle sensors are used for heavy computation, then measurement precision and reliability are improved, but device complexity and cost increase
Solution Approach 1:
The patent combines multiple functions (optical flow computation, vanishing point detection, vehicle detection) into a single integrated processing unit within the camera module. This merging eliminates the need for separate external image processors and reduces system complexity while maintaining detection accuracy through coordinated processing of multiple functions in one unit.
Solution Approach 2:
The processing unit within the camera is designed to perform multiple functions: computing optical flow, detecting vanishing points, identifying vehicles, and determining traffic conditions. This multi-functional approach replaces the need for dedicated separate sensors and processors, reducing device complexity while preserving measurement precision through unified processing.
2Productivity
If external image processors are used for heavy computation, then processing capability is improved, but ease of manufacture and cost worsen
Solution Approach 1:
The patent integrates the image processing unit directly into the camera module, combining what would traditionally be separate components (camera sensor and external processor). This integration reduces the number of parts that need to be manufactured and assembled, lowering manufacturing costs while maintaining high processing capability for optical flow and vehicle detection.
Solution Approach 2:
The camera module becomes self-sufficient by incorporating its own processing unit that can independently compute optical flow, detect vanishing points, and identify vehicles without requiring external processing assistance. This self-service capability eliminates the need for additional external processors, reducing manufacturing complexity and cost.
3Measurement precision
If additional sensors are added for better detection, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent enhances detection precision by changing the processing parameters and algorithms within the existing camera system rather than adding more sensors. By implementing advanced optical flow computation, vanishing point detection, and motion vector analysis, the system achieves high vehicle detection accuracy using only the camera's native sensing capabilities, avoiding additional sensor complexity.
4Measurement precision
If heavy computation is performed externally, then detection accuracy is improved, but power consumption increases
Solution Approach 1:
The camera module performs all heavy computation tasks (optical flow, vanishing point detection, vehicle identification) internally within its own integrated processing unit, eliminating the need to transmit large amounts of raw image data to external processors. This self-service approach reduces power consumption by processing data locally rather than requiring continuous high-bandwidth communication and external processing power.
Data Source
AI summary
A sequence of images obtained by a camera mounted on a vehicle is processed in order to generate Optical Flow data including a list of Motion Vectors being associated with respective features in the sequence of images. The Optical Flow data is analyzed to calculate a Vanishing Point by calculating the mean point of all intersections of straight lines passing through motion vectors lying in a road. An Horizontal Filter subset is determined taking into account the Vanishing Point and a Bound Box list from a previous frame in order to filter from the Optical Flow the horizontal motion vectors. The subset of Optical Flow is clustered to generate the Bound Box list retrieving the moving objects in a scene. The Bound Box list is sent to an Alert Generation device and an output video shows the input scene where the detected moving objects are surrounded by a Bounding Box.