Point Cloud Encoding Segmentation for Latency Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in efficiently processing and compressing large amounts of point cloud data, particularly for applications like virtual reality, augmented reality, and self-driving services, due to high latency and complex encoding/decoding processes.
Innovation Solution
A method and apparatus for transmitting and receiving point cloud data that involves acquiring data through LiDAR equipment, encoding it by splitting points into road and object categories based on radius information, and transmitting the encoded data with signaling information to improve compression and reduce latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If point cloud data is processed using traditional encoding methods, then the data can be transmitted, but the encoding/decoding complexity is high and latency increases
Solution Approach 1:
The patent segments point cloud data into multiple types (static, dynamic, foreground, background) based on motion detection and clustering analysis. This segmentation allows different encoding strategies to be applied to different segments, reducing overall encoding complexity and enabling parallel processing, which reduces latency.
Solution Approach 2:
The patent applies different encoding qualities and compression ratios to different types of point cloud segments. High-priority segments (e.g., foreground objects) are encoded with higher quality, while low-priority segments (e.g., background) use lower quality encoding. This localized quality approach reduces total encoding complexity and time while maintaining perceptual quality.
2Ease of manufacture
If all point cloud data is compressed uniformly, then the process is simple, but compression performance is insufficient for diverse point cloud content
Solution Approach 1:
The patent implements dynamic classification of point cloud data into multiple types based on motion characteristics and spatial distribution. The system adapts encoding parameters dynamically according to the detected point cloud type, achieving both simplified automated processing and optimized compression performance for diverse content.
Solution Approach 2:
The patent changes encoding parameters (compression ratio, quantization steps, transformation methods) based on the detected point cloud type. Different parameter sets are applied to different segments, enabling optimal compression performance for each type while maintaining overall process automation and simplicity.
3Productivity
If motion vectors are applied to all prediction units, then compression efficiency improves, but encoding complexity increases
Solution Approach 1:
The patent applies motion vector compensation selectively only to certain prediction units that benefit most from it, rather than to all prediction units. This partial application maintains compression efficiency for critical segments while reducing overall encoding complexity and computational burden.
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 efficient compression and transmission of point cloud data, reducing latency and improving compression performance, thereby supporting high-quality point cloud services and applications like autonomous driving.
Implementation Method 1
acquiring point cloud data including points through LiDAR equipment having laser sensors
Data Source
AI summary
Disclosed are a method for transmitting point cloud data, a device for transmitting cloud data, a method for receiving cloud data, and a device for receiving cloud data according to embodiments. The method for transmitting point cloud data according to embodiments may comprise the steps of: obtaining point cloud data including points through lidar equipment equipped with laser sensors; encoding the point cloud data; and transmitting the encoded point cloud data and signaling data.