Equi-Height Mosaicking for Vehicle Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle operation assisting systems face significant delays in image processing and object detection due to the complexity of processing images captured by cameras, which are essential for immediate obstacle detection during vehicle operations.
Innovation Solution
The proposed solution involves an image processing device and method that extracts areas corresponding to target objects from images, creates a mosaicking image by connecting these areas, and detects objects within the mosaicking image, utilizing a vanishing point detection unit, area height calculation, and object detection unit to reduce processing time and calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system processes the entire captured image to detect objects, then the detection accuracy is improved, but the processing time increases significantly
Solution Approach 1:
The patent divides the image processing task into two stages: first, extract only the regions containing target objects (vehicles, pedestrians) using size-based filtering; second, perform object detection only on these extracted regions. This segmentation approach maintains detection accuracy while significantly reducing the amount of data requiring complex processing.
Solution Approach 2:
The patent extracts and isolates only the relevant portions of the image that contain objects of interest (vehicles, pedestrians) based on predetermined size criteria. By taking out only these necessary regions and discarding the rest, the system reduces processing time while maintaining the ability to detect all relevant objects accurately.
2Reliability
If the system processes the entire captured image to detect objects, then all objects are detected completely, but the calculation complexity increases
Solution Approach 1:
The patent segments the image processing workflow into a preliminary extraction phase and a detection phase. The extraction phase uses simple size-based filtering to identify regions containing vehicles or pedestrians, separating this from the more complex object detection phase. This reduces overall calculation complexity while maintaining detection completeness.
Solution Approach 2:
The patent performs preliminary action by extracting and preparing only the regions containing target objects before performing the main object detection task. This preliminary size-based filtering reduces the data volume and complexity of subsequent detection calculations while ensuring all relevant objects are included.
3Measurement precision
If the system uses detailed image processing to detect objects accurately, then the detection precision is improved, but the processing speed decreases
Solution Approach 1:
The patent segments the processing task to apply different levels of analysis to different parts of the image. Simple size-based extraction is applied to the entire image, while detailed object detection algorithms are applied only to the extracted regions containing potential objects. This maintains detection precision while improving processing speed through selective detailed analysis.
Data Source
AI summary
Provided is an image processing device including: an area extraction unit extracting areas corresponding to size of a predetermined target object from an image; a mosaicking image creation unit connecting extracted areas to create a mosaicking image; and an object detection unit detecting an object from the mosaicking image.