Vehicle Object Detection Using Balanced Image Reference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection systems struggle to accurately detect the position of objects in front of a vehicle, especially when the vehicle is pitching due to passengers or cargo, as they rely on variations in y-coordinate over time, which are difficult to correct for.
Innovation Solution
An apparatus and method that determine whether an image was captured when the vehicle was balanced, using a first image taken when the vehicle was balanced to correct the position of an object in a second image, even if the vehicle was pitching, by computing the y-coordinate, width, and distance based on camera parameters and optical flow analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple objects are extracted from images and y-coordinate correction is performed based on variation over time, then the position of objects can be detected by excluding the influence of vehicle pitching, but it becomes difficult to detect variation in object position due to pitching caused by passengers, cargo, etc.
Solution Approach 1:
The system performs preliminary identification of balanced-state images before processing object position detection. By pre-selecting images captured when the vehicle is in a balanced state (where the camera optical axis is substantially horizontal), the system establishes a reliable reference frame for subsequent position measurements, eliminating the need to correct for pitching variations during normal operation.
Solution Approach 2:
The system uses images captured when the vehicle is in a balanced state as reference copies for correcting object positions in images taken during pitching. By comparing and mapping object positions between the reference balanced-state images and current pitching-state images, the system accurately determines true object positions independent of vehicle attitude variations.
2Measurement precision
If y-coordinate detection is performed based on variation over time of object position, then correction for vehicle pitching can be achieved, but additional sensors and complex correction algorithms are required
Solution Approach 1:
The system uses the image pickup device itself to determine vehicle balance state by analyzing image characteristics, eliminating the need for external balance sensors. The controller identifies balanced-state images by detecting when the camera optical axis is substantially horizontal based solely on image data, making the system self-sufficient and reducing overall complexity.
Solution Approach 2:
The image pickup device serves multiple functions: it captures object images for position detection and simultaneously provides data for determining vehicle balance state. This multi-functionality eliminates the need for separate sensors for balance detection, reducing device complexity while maintaining measurement precision.
3Productivity
If object position is tracked using continuous image sequences, then smooth tracking can be achieved, but accuracy deteriorates when the vehicle is pitching
Solution Approach 1:
The system implements feedback by continuously comparing current object positions with positions derived from balanced-state reference images. When pitching occurs, the controller uses the pre-established reference data to correct position measurements, ensuring accurate tracking continues even during vehicle attitude changes.
Solution Approach 2:
The system performs preliminary capture and storage of balanced-state images before pitching occurs. These pre-captured reference images serve as the basis for correcting position measurements during subsequent pitching events, allowing continuous accurate tracking without requiring real-time balance detection during each measurement.
Data Source
AI summary
An apparatus for detecting an object is provided which detects the pitching of a vehicle, and detects the presence of an object the front of the vehicle itself with the pitching being taken into consideration. The apparatus includes a microcomputer that performs image processing for the image of the object to compute the velocity and acceleration of the object present in the picked-up image, and based on the computed acceleration of the object, to determine whether the image was captured when the vehicle itself was balanced. If the image is judged to be an image captured when the vehicle itself was not balanced, then the position of the object present in the picked-up image is computed based on another picked-up image that was captured when the vehicle itself was balanced.