LiDAR Dynamic Field-of-View Resolution Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional LiDAR systems face challenges in scanning large fields-of-view (FOVs) at high resolutions due to limitations in detector arrays and laser power, leading to reduced detection accuracy and range, making it difficult to achieve point clouds with high resolution and large FOVs simultaneously.
Innovation Solution
A LiDAR system that dynamically selects a first FOV for low-resolution scanning and a second FOV for high-resolution scanning within the first FOV, using separate transmitter subsystems and a controller to identify and adjust the sizes and locations of these FOVs based on object data, allowing for improved object identification and navigation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of pixels in the photodetector array is increased to improve point-cloud resolution, then measurement precision is improved, but the laser power received by each pixel decreases, worsening detection accuracy
Solution Approach 1:
The patent applies local quality by implementing a multi-resolution scanning strategy where different regions of the field-of-view are scanned at different resolutions. Specifically, a first FOV is scanned at low resolution while a second FOV containing the area-of-interest is scanned at high resolution. This allows high measurement precision to be concentrated in the region where it is most needed (the area-of-interest) while maintaining acceptable precision elsewhere, thereby preserving sufficient laser power per pixel in the high-resolution region without requiring a uniformly high pixel count across the entire detector array.
2Measurement precision
If point-cloud resolution is increased from 0.1° to 0.01°, then measurement precision is improved, but the number of pixels required increases by a factor of one-hundred, worsening device complexity
Solution Approach 1:
The patent applies segmentation by dividing the field-of-view into multiple regions: a first FOV that is scanned at low resolution and a second FOV (within the first FOV) that is scanned at high resolution. This segmentation allows the system to achieve high point-cloud resolution (0.01°) only in the second FOV where the area-of-interest is located, while using lower resolution (e.g., 0.1°) in the rest of the field-of-view. Consequently, the total number of pixels required is dramatically reduced compared to a uniform high-resolution system, as the high-resolution pixel array is only needed for a portion of the total scanning area.
3Measurement precision
If point-cloud resolution is increased, then measurement precision is improved, but detection range decreases, worsening the system's ability to sense distant objects
Solution Approach 1:
The patent applies local quality by implementing differential resolution scanning where a first FOV is scanned at low resolution and a second FOV containing the area-of-interest is scanned at high resolution. This allows the system to maintain high detection range for the overall field-of-view using low-resolution scanning, while achieving high measurement precision (fine resolution) localized to the area-of-interest. The low-resolution scanning of the first FOV preserves detection capability at long distances, while the high-resolution scanning of the second FOV provides detailed measurement precision where needed, thus resolving the contradiction between resolution and detection range.
4Productivity
If the scanning period is reduced to 100 milliseconds for the entire FOV, then productivity is improved, but the ability to scan high-resolution point clouds is worsened
Solution Approach 1:
The patent applies segmentation by dividing the scanning task into two parts: scanning a first FOV at low resolution and scanning a second FOV (within the first FOV) at high resolution. This segmentation enables the system to complete the low-resolution scan of the entire FOV within the 100-millisecond scanning period, maintaining high productivity. The high-resolution scanning is then performed on the smaller second FOV, which requires less time and can be completed within the remaining time budget or in subsequent scanning cycles, thereby achieving high measurement precision without compromising the overall scanning speed and productivity.
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 enables higher accuracy in object identification and autonomous navigation by selectively using fine resolution only where necessary, maintaining reasonable photodetector size and laser power for long-distance, high-resolution point-cloud generation.
Implementation Method 1
a typical LiDAR system measures the distance to a target by illuminating the target with pulsed laser light beams that are steered towards an object in the far field using a scanning mirror, and then measuring the reflected pulses with a sensor
Implementation Method 2
a typical LiDAR system measures the distance to a target by illuminating the target with pulsed laser light beams that are steered towards an object in the far field using a scanning mirror, and then measuring the reflected pulses with a sensor
Implementation Method 3
measuring the reflected pulses with a sensor
Implementation Method 4
Differences in laser light return times, wavelengths, and/or phases (also referred to as 'time-of-flight (ToF) measurements') can then be used to construct digital three-dimensional (3D) representations of the target
Data Source
AI summary
Embodiments of the disclosure provide for a LiDAR system. The LiDAR system may dynamically select a first FOV of a far-field environment to be scanned at a rough resolution and a second FOV including important information, as indicated based on object data from a previous scanning procedure, to be scanned at a fine resolution. For example, an area-of-interest, such as along the horizon where pedestrians, vehicles, or other objects may be located, may be scanned with the finer resolution. Using fine resolution for the area-of-interest may achieve a higher-degree of accuracy/safety in terms of autonomous navigation decision-making than if coarse resolution is used. Because the use of fine resolution is limited to a relatively small area, a reasonably sized photodetector and laser power may still be used to generate a long distance, high-resolution point-cloud.