LiDAR Object Detection Using Ground Data to Resolve Low Reflectivity Gaps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
LiDAR sensors face challenges in accurately detecting objects with parts difficult to reflect light, leading to separation of objects and reduced recognition accuracy due to incorrect recognition of undetected areas as part of larger objects.
Innovation Solution
An object detection apparatus with a sorting unit to categorize distance measurement data, a determination unit to assess the presence of the ground in undetected areas, and an estimation unit to identify obstacles straddling adjacent areas, improving object detection accuracy by preventing separation of objects with low light reflection properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LiDAR sensor is used to detect objects by measuring scattered light, then distance measurement capability is improved, but detection accuracy deteriorates when objects have parts difficult to reflect light
Solution Approach 1:
The patent introduces ground information as an intermediary element to infer the presence of objects. When light does not return from a specific direction, the system checks whether the ground exists in that direction using ground data. This intermediary approach allows the system to distinguish between objects with low reflectivity and actual absence of objects, thereby improving detection reliability while maintaining distance measurement capability.
Solution Approach 2:
The patent creates a virtual representation of the ground environment by storing ground information in advance. This copied ground data serves as a reference to determine whether undetected areas are due to object presence (with low reflectivity) or actual absence of objects, enabling accurate object detection even when direct light reflection is insufficient.
2Stability of the object's composition
If undetected points are filled with information from adjacent points, then object recognition continuity is improved, but object recognition accuracy deteriorates due to incorrect merging of multiple objects
Solution Approach 1:
The patent applies different processing rules to different spatial relationships. Instead of uniformly filling all undetected points with adjacent point information, the system specifically uses ground information to determine whether to infer object presence in undetected areas. This localized approach maintains object recognition continuity where appropriate while preventing incorrect merging of multiple objects by relying on ground-based spatial reasoning.
3Reliability
If ground information is used to determine undetected areas, then object detection reliability is improved, but device complexity increases
Solution Approach 1:
The patent performs preliminary action by storing ground information in advance before actual object detection. The ground data is prepared and stored beforehand, allowing the detection system to quickly query this pre-processed information when encountering undetected areas. This preliminary preparation reduces the computational burden during real-time detection while improving reliability through ground-based spatial reasoning.
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 object detection accuracy by correctly identifying obstacles even when they have parts difficult to reflect light, reducing incorrect recognition and separation of objects.
Implementation Method 1
a sensor (51) that measures a distance by observing reflected waves of a plurality of signals (laser beams) radiated toward areas different from one another
Data Source
AI summary
In an object detection apparatus (10), a sorting unit (20) obtains data indicating a measurement result of a distance of each signal that a sensor (51) radiated from the sensor (51). The sorting unit (20) sorts the data obtained into first data indicating that the distance was not measured, second data indicating a distance to one or more obstacles, and third data indicating a distance to the ground. A determination unit (30), by referring to the third data, determines whether or not the ground exists in a first area which is a radiation destination area of a signal corresponding to the first data. An estimation unit (40) estimates, when the determination unit (30) determines that the ground exists in the first area, and when the first area and a second area, which is a radiation destination area of a signal corresponding to the second data, are adjacent, that one obstacle exists straddling the first area and the second area.