Object Recognition Device Using Color Image Interpolation for LIDAR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object recognition devices using LIDAR struggle to accurately recognize low-reflecting objects due to absorption of laser beams, leading to incomplete data sets and incorrect clustering processing.
Innovation Solution
An object recognition device that incorporates a LIDAR system and a processing unit capable of executing clustering processing based on 3D positional data, including time of flight data, and utilizes color image information to associate detection points across undetected coordinates, employing interpolation to fill gaps in data sets, particularly using time of flight data from nearby coordinates or road surface data for accurate clustering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LIDAR is used to detect objects by measuring time of flight of laser beam, then 3D positional data can be obtained, but low reflecting objects that absorb laser beam cannot be detected resulting in blank regions in data point group
Solution Approach 1:
The patent introduces color image data as an intermediary to bridge the detection gap for low-reflecting objects. When LIDAR fails to detect an object (creating a blank region), the system uses color camera data from the corresponding spatial location to infer the presence and properties of the undetected object, thereby maintaining detection reliability without compromising measurement precision
Solution Approach 2:
The system performs preliminary detection using multiple sensors (LIDAR and color camera) simultaneously. By capturing data from both sensors before clustering processing, the system prepares complementary information sets that can be cross-referenced to identify and fill detection gaps, ensuring reliable object recognition even when one sensor fails to detect low-reflecting objects
2Productivity
If clustering processing is executed only on detected points with valid time of flight data, then processing efficiency is maintained, but low reflecting objects are misidentified or missed
Solution Approach 1:
Color image data serves as an intermediary to supplement incomplete LIDAR detection data. The system uses the color image information to identify potential objects in blank regions and associates them with nearby detected points through the linkable condition, enabling accurate object recognition without requiring extensive additional processing of undetected regions
Solution Approach 2:
The patent applies partial action by selectively extending clustering processing only to regions where blank zones are identified through color image analysis. Instead of processing the entire data set with enhanced algorithms, the system focuses computational resources on specific areas where low-reflecting objects are suspected, maintaining overall processing efficiency while improving recognition accuracy
3Device complexity
If blank regions are excluded from clustering processing, then processing complexity is reduced, but recognition of low reflecting objects becomes erroneous
Solution Approach 1:
The patent uses color image data as an intermediary to identify and mark blank regions that likely contain low-reflecting objects. This intermediary information allows the system to selectively apply enhanced clustering algorithms only to specific blank regions rather than the entire data set, maintaining processing simplicity while improving recognition reliability for problematic areas
Solution Approach 2:
The system segments the data processing into two distinct pathways: standard clustering for regions with valid LIDAR data, and enhanced clustering for blank regions identified through color image analysis. This segmentation allows the system to maintain low processing complexity for the majority of data while applying more sophisticated algorithms only where necessary to improve recognition reliability
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
Enhances the accuracy of object recognition by effectively grouping low-reflecting objects into clusters, reducing misidentification and computational load through intelligent data interpolation and association techniques.
Implementation Method 1
The LIDER uses time of flight (TOF) of laser beam to generate the Z positional data (i.e., TOF data)
Data Source
AI summary
In step S11, a color image CIMG is acquired. In step S12, a distance image DIMG is acquired. In step S13, the color image CIMG is projected onto the distance image DIMG. An alignment of the color image CIMG and the distance image DIMG is performed prior to the projection of the color image CIMG. In step S14, it is determined whether or not a basic condition is satisfied. In S15, it is determined whether or not a special condition is satisfied. If a judgement result of the steps S14 or S15 is positive, then in step S16 a first data point and a second data point on the distance image DIMG are associated. If bot of the judgement results of steps S14 and S15 are negative, data points are not associated in step S17.