Image Processing Apparatus for Far Object Detection via LiDAR Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
It is challenging to detect far objects in images due to overlap with near objects, making it difficult to accurately recognize distant objects from two-dimensional data.
Innovation Solution
An image processing apparatus that acquires both two-dimensional and three-dimensional data, extracts a region of interest with valid distance information, and cuts out an object image from the two-dimensional data to enhance detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only two-dimensional image data is used for object detection, then the system complexity is low, but the detection precision of far objects deteriorates due to overlap with near objects
Solution Approach 1:
The patent combines two-dimensional image data from a camera with three-dimensional distance data from LiDAR to create a composite detection system. The 2D image provides visual information while the 3D point cloud data provides depth information, and their integration allows for accurate far object detection by eliminating overlap issues through depth-based segmentation.
Solution Approach 2:
The patent transitions from two-dimensional image data to three-dimensional spatial data by incorporating LiDAR point cloud information. This dimensional upgrade enables the system to distinguish between near and far objects along the depth axis, resolving the overlap problem that plagues 2D detection systems.
2Measurement precision
If three-dimensional data is integrated with two-dimensional image data, then the detection precision of far objects improves, but the device complexity increases
Solution Approach 1:
The patent segments the detection process into distinct functional modules: a 2D data acquisition unit for image capture, a 3D data acquisition unit for LiDAR data collection, a region setting unit for defining search areas, and an extraction unit for isolating target objects. This modular segmentation manages system complexity by organizing the multi-source data integration into manageable, independent components.
Solution Approach 2:
The patent introduces a region setting unit as an intermediary component that processes both 2D image data and 3D point cloud data to define extraction target regions. This mediator coordinates the integration of multi-dimensional data, managing the complexity of combining different data types while enabling accurate far object detection.
3Reliability
If the entire captured image is processed, then comprehensive object detection is achieved, but the information processing complexity increases
Solution Approach 1:
The patent applies local quality by setting specific extraction target regions within the overall image based on 3D distance information. Instead of uniformly processing the entire image, the system identifies and focuses computational resources on specific regions where far objects are likely to be located, thereby reducing processing complexity while maintaining detection reliability.
Solution Approach 2:
The patent implements partial action by extracting and processing only the relevant portions of the image that contain potential far objects, rather than analyzing the entire captured image. The region setting unit defines these partial regions based on 3D spatial data, reducing the overall processing load while ensuring that all potential targets are covered.
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 accurate detection of distant objects by isolating the object of interest from overlapping near objects, improving recognition accuracy and reducing information processing complexity.
Implementation Method 1
three-dimensional data acquisition unit that acquires three-dimensional data for at least a partial region of a range captured as the image
Data Source
AI summary
An image processing apparatus according to an example embodiment includes: a two-dimensional data acquisition unit that acquires an image that is two-dimensional data; a three-dimensional data acquisition unit that acquires three-dimensional data for at least a partial region of a range captured as the image; an extraction target region setting unit that outputs two-dimensional coordinates of a region including point cloud data in which distance information of the three-dimensional data is within a preset recognition section as extraction target region coordinates; and an object image extraction unit that extracts an image of the region corresponding to the extraction target region coordinates from the image as an object image.