Lidar Data Slicing for Low-Latency 360° Scan Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The latency in LIDAR data processing pipelines, particularly in systems with limited compute resources, leads to resource contention and visual artifacts due to the need for full revolutions of data before preprocessing can begin, causing peaks in compute needs and suboptimal processing latency in camera/LIDAR fusion operations.
Innovation Solution
Processing LIDAR data in streams or slices within a full revolution, rather than waiting for a full revolution to complete, allows for more even compute loads and reduces latency by publishing preprocessed data for smaller field-of-views, thereby eliminating idle wait times and visual artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LIDAR waits for full revolution to complete before preprocessing data, then full field-of-view coverage is achieved, but processing latency increases and resource contention occurs
Solution Approach 1:
The patent divides the full revolution LIDAR data into multiple partial scans or segments that can be preprocessed independently and in parallel. This segmentation allows different portions of the rotational data to be processed simultaneously rather than waiting for the complete 360-degree scan, thereby reducing overall processing latency while maintaining comprehensive field-of-view coverage through aggregation of the segmented results.
2Loss of information
If LIDAR processes data after full revolution completion, then complete scene coverage is obtained, but compute resource contention increases due to peak processing demands
Solution Approach 1:
The patent performs preliminary preprocessing operations on LIDAR data segments as they become available during the rotational scan, rather than waiting for complete data acquisition. This preliminary action includes initial filtering, validation, and organization of data subsets, which distributes computational workload throughout the scanning process and prevents resource contention peaks that would occur if all processing were deferred until after the full revolution completed.
3Quantity of substance
If LIDAR uses full revolution scans, then comprehensive data collection is achieved, but visual artifacts appear due to latency in camera-LIDAR fusion
Solution Approach 1:
The patent implements continuous preprocessing of LIDAR data segments as they are acquired during rotation, maintaining an ongoing processing pipeline rather than batch processing after completion. This continuous action ensures that processed LIDAR data is available in near-real-time for fusion with camera feeds, eliminating temporal misalignment and visual artifacts that arise from latency when using traditional full-revolution-wait approaches.
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 reduces resource contention and latency, improves the availability of LIDAR data, and eliminates visual artifacts by processing LIDAR data in a semi-continuous frequency, allowing for earlier utilization of partial LIDAR data in camera-LIDAR fusion operations.
Implementation Method 1
a light ranging and detection (LIDAR) sensor can be used to determine ranges (variable distance) of one or more targets by directing a laser to a surface of an entity (e.g., a person, an object, a structure, an animal, etc.) and measuring the time for light reflected from the surface to return to the LIDAR
Data Source
AI summary
Systems and techniques are provided for processing data from an optical sensor. An example method includes obtaining, from an optical sensor configured to rotate about an axis, sensor data; generating, based on the sensor data, slices of sensor data, each slice having a field-of-coverage (FOC) that is less than 360 degrees, wherein a slice size is determined based on a rate for publishing a combination of slices that yields 360 degrees of coverage within a threshold period, a number and size of slices estimated to yield a combined FOC of 360 degrees while achieving a desired reduction in a resources contention by consumer nodes, and/or a field-of-view (FOV) of a camera device; and providing, to the consumer nodes, a partial optical sensor scan comprising the slices of sensor data, the partial optical sensor scan being provided prior to obtaining a revolution of sensor data having a 360 degrees of coverage.