Image Sensor Data Prioritization for Dynamic Point Cloud Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The acquisition of three-dimensional information in real-time using LiDAR results in a large data volume, necessitating efficient data management by prioritizing image sensor data for static and dynamic objects in real spaces, which is also applicable to other sensors acquiring point cloud data.
Innovation Solution
An image sensor data control system comprising terminal devices and a server device connected for data communication, where the server learns motion feature indices of image sensor data to prioritize data transmission based on static or dynamic objects, with higher priority given to dynamic objects for real-time data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LiDAR is used to acquire three-dimensional information in real-time, then measurement precision and reliability are improved, but the quantity of data increases enormously
Solution Approach 1:
The patent segments the acquired point cloud data into multiple spatial regions and further divides objects into static and dynamic categories. By processing different regions and object types separately, the system manages the enormous data volume through structured segmentation while preserving the high measurement precision of LiDAR for all categories.
Solution Approach 2:
The patent extracts and prioritizes data for dynamic objects from the complete point cloud dataset, separating it from static object data. This extraction allows the system to focus computational and transmission resources on the most critical moving elements, reducing the effective data volume that requires real-time processing while maintaining accuracy for dynamic targets.
2Measurement precision
If multiple LiDAR systems are used to improve the quality of three-dimensional information, then measurement precision and reliability are improved, but the data volume increases further
Solution Approach 1:
The patent merges point cloud data from multiple LiDAR systems into a unified spatial representation. By combining data from multiple sensors and integrating it with image data, the system achieves comprehensive coverage and high point density while managing the total data volume through coordinated processing of the merged dataset rather than handling each LiDAR's output separately.
Solution Approach 2:
The patent applies different processing qualities and priorities to different spatial regions based on their importance. High-priority regions containing dynamic objects receive more detailed processing and higher point density, while low-priority static regions use reduced processing. This local quality differentiation maintains measurement precision where needed while reducing overall data volume from multiple LiDAR systems.
3Reliability
If all image sensor data is transmitted in real-time, then reliability of information delivery is improved, but loss of time for data transmission increases
Solution Approach 1:
The patent implements dynamic prioritization of data transmission based on object characteristics and spatial region importance. Transmission parameters such as frequency, resolution, and timing are adjusted dynamically according to whether objects are static or dynamic and their location in the environment. This dynamic approach ensures reliable delivery of critical dynamic object data while reducing transmission time for static regions.
Solution Approach 2:
The patent applies partial action by selectively transmitting only the most critical data subsets in real-time, specifically focusing on dynamic objects and high-priority regions. Rather than transmitting all image sensor data with equal priority, the system performs partial transmission of essential data, ensuring reliability for moving objects while minimizing overall transmission time through selective data submission.
Data Source
AI summary
A terminal device acquires image sensor data in real space from a sensor unit and transmits the data to a server device by a transmission unit. The server device aggregates the image sensor data in the real space transmitted from each terminal device by the aggregation unit, and then learns about the motion feature index of the image sensor data composed of point clouds in each spatial region in the real space by the learning unit to determine the spatial region pertaining to a dynamic object or static object in the real space, and stores it in a motion feature index information storage unit. The server device preferentially transmits the image sensor data in the spatial region determined to be a dynamic object by the learning unit to the moving object, with respect to the image sensor data acquired in real time.