Remote Support Image Projection for Moving Object Delay Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing remote vehicle monitoring systems suffer from inaccuracies due to delays in processing camera images and object recognition results, leading to significant deviations in the position and size of moving objects in synthetic images.
Innovation Solution
A system that performs projection transformation on partial images of moving bodies to predict their future positions, adjusting the images based on predicted movement amounts and generating synthetic images with superimposed annotation information to minimize deviations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If camera images and object recognition results are transmitted separately from the mobile vehicle to the remote server, then the system can process and display remote vehicle monitoring information, but the delay in processing causes significant deviation in position and size of moving objects in the synthetic image
Solution Approach 1:
The system performs preliminary actions by predicting the movement amount of moving bodies before generating the synthetic image. The remote server calculates predicted movement amounts based on the time delay between image capture and recognition result transmission, then pre-adjusts the position and size of moving bodies in the synthetic image to compensate for the delay, ensuring accurate representation of current positions.
Solution Approach 2:
The system applies preliminary anti-action by introducing a counteracting adjustment to offset the harmful effect of processing delay. The remote server calculates the deviation that would occur due to delay time and applies an equal and opposite adjustment to the moving body positions and sizes in the synthetic image, thereby canceling out the error caused by the delay.
2Ease of operation
If the remote server processes object recognition results based on images captured at different timings, then the system can generate synthetic images with annotation information, but the position and size of moving bodies greatly deviate from actual positions and sizes
Solution Approach 1:
The system changes parameters by dynamically adjusting the position and size parameters of moving bodies in the synthetic image based on predicted movement amounts. The remote server modifies these parameters to compensate for the time delay, transforming the inaccurate representation into an accurate one that reflects the current state of moving bodies.
Solution Approach 2:
The system implements feedback by using the time delay information and movement prediction results to continuously adjust the position and size parameters of moving bodies in the synthetic image. This closed-loop approach ensures that the synthetic image accurately represents the current state despite the processing delay.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
When an object to be noted in an image IMG1 is a moving body, projection transformation of a partial image OBJ1 of the moving body included in the image IMG1 used for a recognition of the moving body is performed. According to the projection transformation of the partial image OBJ1, a partial image OBJ2 obtained at a camera viewpoint ahead of an acquisition timing T1 of the image IMG1 including the partial image OBJ1 by an object movement adjustment time α is generated. A synthetic image SIMG1 is generated based on the partial image OBJ2, the image IMG1 in which the partial image OBJ1 is included and recognition information OR of the moving body in the image IMG1, and is output from a display device.