Point Cloud Coding via Classification-Based Motion Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional point cloud coding techniques face inefficiencies in coding accuracy and complexity due to inaccurate classification of points, high complexity in the classification process, and wasteful processing of rotation matrices, especially in sparse Lidar point cloud data captured by moving vehicles.
Innovation Solution
The proposed method employs classification-based global motion estimation using multiple thresholds and types of process units to accurately classify points into road and object categories, and utilizes fixed or multiple rotation matrices for motion compensation to improve coding efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional classification methods are used to classify points in point cloud data, then the classification process can be performed, but the classification accuracy is low and the complexity is high
Solution Approach 1:
The patent segments the point cloud data into different types of process units (e.g., road points, object points, background points) based on spatial distribution and motion characteristics. This segmentation enables more accurate classification by treating different regions with appropriate methods, reducing overall classification complexity through divide-and-conquer strategy
Solution Approach 2:
The patent applies different classification strategies to different local regions of the point cloud data. For example, road points are classified using horizontal motion characteristics while object points use vertical motion characteristics. This local quality approach improves classification accuracy by matching the classification method to the specific characteristics of each region
2Measurement precision
If multiple rotation matrices are used for motion compensation, then the accuracy of motion compensation is improved, but the computational complexity and processing resources increase
Solution Approach 1:
The patent implements a dynamic rotation matrix selection mechanism where the system adaptively chooses between fixed rotation matrices and multiple rotation matrices based on the specific characteristics of the point cloud data and motion patterns. This dynamic approach maintains high accuracy when needed while reducing complexity when simple motion is sufficient
Solution Approach 2:
The patent changes the parameter of rotation matrices from fixed to adaptive, allowing the system to select appropriate rotation matrices based on motion characteristics. This parameter change enables the system to achieve high motion compensation accuracy for complex motions while using simpler fixed matrices for straightforward cases, balancing accuracy and complexity
3Productivity
If fixed rotation matrix is used for motion compensation, then the processing resources and computational complexity are reduced, but the accuracy of motion compensation may be insufficient for complex motions
Solution Approach 1:
The patent implements a dynamic rotation matrix selection mechanism where the system adaptively chooses between fixed rotation matrices and multiple rotation matrices based on the specific characteristics of the point cloud data and motion patterns. This dynamic approach maintains high accuracy when needed while reducing complexity when simple motion is sufficient
Solution Approach 2:
The patent changes the parameter of rotation matrices from fixed to adaptive, allowing the system to select appropriate rotation matrices based on motion characteristics. This parameter change enables the system to achieve high motion compensation accuracy for complex motions while using simpler fixed matrices for straightforward cases, balancing accuracy and complexity
4Productivity
If conventional point cloud coding techniques are used, then the coding process can be performed, but the coding efficiency is insufficient
Solution Approach 1:
The patent segments the point cloud data into different types of process units (e.g., road points, object points, background points) based on spatial distribution and motion characteristics. This segmentation enables more accurate classification by treating different regions with appropriate methods, reducing overall classification complexity through divide-and-conquer strategy
Solution Approach 2:
The patent applies different classification strategies to different local regions of the point cloud data. For example, road points are classified using horizontal motion characteristics while object points use vertical motion characteristics. This local quality approach improves classification accuracy by matching the classification method to the specific characteristics of each region
Data Source
AI summary
Embodiments of the present disclosure provide a method for point cloud coding. The method comprises: obtaining, during a conversion between a current frame of a point cloud sequence and a bitstream of the point cloud sequence, a compensated reference frame of the current frame by performing motion compensation on a reference frame of the current frame based on a set of rotation matrixes comprising at least one fixed rotation matrix or a plurality of rotation matrixes; and performing the convention based on the compensated reference frame. Compared with the conventional solution, the proposed method can advantageously improve coding efficiency.