Point Cloud Compression for Direct Processing in Autonomous Vehicles

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional technologies for compressing point cloud data in autonomous vehicles require decompression before mathematical operations, which is inefficient and limited by memory management issues, especially in vehicles with smaller and cheaper memory devices.

Innovation Solution

The method involves splitting 3D space into tiles to limit numerical ranges for point cloud data representation, using a fixed-point Q-format representation to convert floating-point coordinates into integer representations, allowing for on-the-fly compression and mathematical operations on compressed data without prior decompression.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional compression technologies are used, then point cloud data can be compressed, but decompression is required before mathematical operations which reduces processing efficiency

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidtime for decompression and processing

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent segments the point cloud data into multiple tiles, where each tile is independently compressed and can be processed separately. This segmentation allows mathematical operations to be performed on individual tiles without requiring full decompression, thereby improving processing efficiency and reducing time loss.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent changes the representation parameters of point cloud data by encoding coordinates in a compressed format that preserves mathematical operability. By transforming the data representation while maintaining its essential geometric properties, the system enables direct mathematical operations on compressed data, eliminating the need for decompression and thus improving productivity.

Inventive Principle:

Principle #35Parameter changes

2Quantity of substance

If memory device size is increased to store more point cloud data, then storage capacity improves, but memory fragmentation and virtualization issues arise

Engineering Contradiction:

Improvestorage capacityVSAvoidmemory management complexity

Core Design Contradiction:

Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent divides the point cloud data into multiple smaller tiles that can be stored and managed independently. This segmentation reduces the memory management burden by allowing the system to handle smaller, more manageable data units, thereby improving storage capacity utilization while reducing memory fragmentation and virtualization complexity.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent implements a mechanism where less frequently accessed point cloud data can be discarded from active memory and recovered when needed. This approach optimizes storage capacity by dynamically managing memory resources, allowing the system to store more data effectively while reducing the complexity of permanent memory allocation and management.

Inventive Principle:

Principle #34Discarding and recovering

3Quantity of substance

If point cloud data is compressed, then storage efficiency improves, but conventional methods require decompression which increases bandwidth usage

Engineering Contradiction:

Improvedata storage efficiencyVSAvoidbandwidth usage for decompression

Core Design Contradiction:

Quantity of substanceVSLoss of energy

Solution Approach 1:

The patent transforms the data representation parameters to enable mathematical operations directly on compressed data. By encoding point cloud coordinates in a compressed format that preserves geometric relationships, the system eliminates the need for decompression during processing, thereby improving storage efficiency while reducing bandwidth usage and energy loss.

Inventive Principle:

Principle #35Parameter changes

4Measurement precision

If floating-point representation is used for point cloud coordinates, then numerical precision is maintained, but data size increases reducing compression efficiency

Engineering Contradiction:

Improvecoordinate precisionVSAvoiddata size

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent changes the numerical representation parameters by encoding floating-point coordinates in a compressed integer format that preserves precision. By transforming the coordinate representation to a more efficient format while maintaining the necessary numerical accuracy, the system reduces data size and improves compression efficiency without sacrificing measurement precision.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11367253B2Point cloud data compression in an autonomous vehicle

Publication Date: 2022.06.21 GM GLOBAL TECHNOLOGY OPERATIONS LLC

AI summary

Autonomous vehicles and techniques that can be utilized to compress point cloud data and operate on compressed point cloud data are provided. An autonomous vehicle can include a data compression system can configure point cloud data according to a collection of three-dimensional (3D) tiles representative of the region. Each 3D tile can include a portion of the cloud point data, where each point vector in the portion of the cloud point data can be configured relative to a position vector of the 3D tile defined in a coordinate system of the collection of 3D tiles. The data compression system can utilize a fixed-point Q-format representation based on a defined number of bits to compress at least a portion of the point cloud data. The autonomous vehicle also can include a control system that can operate mathematically on compressed point cloud data, without reliance on prior decompression.