Stereo Camera Parallax Motion Vector Overlap Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surrounding environment recognition devices fail to detect moving three-dimensional objects early when they overlap with other objects at short distances, leading to delayed detection and potential collisions, especially in scenarios like a pedestrian emerging from behind a parked vehicle.
Innovation Solution
The system employs stereo cameras with differing frame rates to generate parallax images, track motion vectors, and analyze their distribution to separate overlapping objects, enabling early detection of moving objects within a three-dimensional area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If three-dimensional object detection is based on distance data from stereo cameras, then the detection method is simple and reliable, but overlapping objects at similar distances cannot be separately detected
Solution Approach 1:
The patent segments the detection process into two independent stages: first detecting three-dimensional objects using distance data from stereo cameras, then separately detecting motion vectors for each detected object. This segmentation allows the system to handle overlapping objects by applying motion-based differentiation after the initial distance-based detection, resolving the contradiction between detection reliability and object separation precision.
Solution Approach 2:
The patent introduces motion vector detection as an intermediary step between distance-based object detection and final object identification. By detecting motion vectors for each detected three-dimensional object, the system can differentiate between overlapping objects that have different motion characteristics, thereby improving object separation precision while maintaining the simplicity and reliability of distance-based detection.
2Measurement precision
If detection relies on apparent separation of overlapping objects, then detection accuracy improves, but detection timing is delayed
Solution Approach 1:
The patent performs preliminary detection of three-dimensional objects using distance data before the objects have visually separated. By immediately following up with motion vector detection on each identified object, the system can differentiate overlapping objects based on their motion characteristics even when they appear merged in the image, thereby eliminating detection delay while maintaining accuracy.
Solution Approach 2:
The patent implements a feedback mechanism where motion vector detection results are used to refine and differentiate the initially detected three-dimensional objects. This feedback loop allows the system to correctly identify and separate overlapping objects based on their different motion patterns, achieving accurate detection without waiting for visual separation.
3Speed
If high frame rate is used for motion vector detection, then detection speed improves, but energy consumption increases
Solution Approach 1:
The patent applies partial action by performing high frame rate motion vector detection only on the specific three-dimensional objects that were detected using distance data, rather than processing the entire image at high frame rate. This selective approach maintains fast detection speed for critical objects while reducing overall energy consumption compared to full high frame rate processing.
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
This approach allows for timely detection of moving objects, even when they overlap with other objects, thereby reducing the risk of collisions by providing accurate and prompt recognition of pedestrians or other hazards.
Implementation Method 1
acquire a pair of images by stereo cameras
Data Source
Figure 1
Figure 2~4
Figure 5~7
AI summary
The purpose of the present invention is to provide a surrounding environment recognition device such that even in a situation where a certain moving three-dimensional object is apparently overlapped with another three-dimensional object, early detection of the certain moving three-dimensional object is enabled. To this end, a surrounding environment recognition device for a moving body is equipped with: imaging units for photographing multiple images in a time series; a three-dimensional object detection unit for detecting three-dimensional objects on the basis of distances of the objects from the imaging units; a vector detection unit for tracking feature points within predetermined areas of the multiple images containing the three-dimensional objects, thereby detecting motion vectors of the feature points; and a moving three-dimensional object detection unit for detecting three-dimensional objects which are present in the areas on the basis of detection results of the vector detection unit.