Remote Vehicle Image Projection for Recognition Synchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing systems for remotely monitoring vehicles face delays in outputting synthetic images due to misalignment in processing times for camera images and object recognition results, leading to inefficiencies in remote support.
Innovation Solution
A technique that involves projection transformation and delay compensation to synchronize the processing of camera images and object recognition data, allowing for the generation of synthetic images ahead of the actual acquisition time, thereby reducing delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If camera images and object recognition results are separately transmitted and processed, then data transmission and processing can be performed in parallel, but the output timing of synthetic images is delayed due to misalignment in processing times
Solution Approach 1:
The system performs projection transformation on the camera image to generate a future image at an earlier timing (acquisition timing minus delay compensation time) before the object recognition result is ready. This preliminary action allows the image processing to be completed in advance, eliminating the waiting time and enabling synchronous output with the object recognition result.
Solution Approach 2:
The system changes the timing parameter of the image by performing projection transformation to create a future image that corresponds to an earlier time point. This parameter transformation (from current acquisition timing to future timing) allows synchronization between image processing and object recognition processing that complete at different times.
2Quantity of substance
If image data is encoded and transmitted to reduce data size, then transmission efficiency is improved, but decoding time adds to the overall processing delay
Solution Approach 1:
The system performs projection transformation and generates the future image before the image data is fully decoded. By initiating image processing operations in advance on the encoded data or partially decoded data, the system eliminates the sequential dependency where decoding must complete before processing can begin, thereby reducing the impact of decoding time on overall latency.
3Reliability
If the system waits for both image data and object recognition result to be fully processed before output, then data synchronization is ensured, but the output delay increases
Solution Approach 1:
The system performs projection transformation to generate a future image in advance, before the object recognition result is fully processed. This preliminary generation of the image component allows the system to prepare one half of the synthetic image data early, reducing the waiting time while maintaining synchronization by matching the future image timing with the object recognition result timing.
Data Source
AI summary
An image delay compensation time α is set to be equal to or less than an image delay time indicating a timing difference D1 between the timing T1 at which an image IMG1 is acquired by a camera CAM and a timing T3 at which a remote operator terminal 200 decodes the image IMG1. Based on information on a movement of a mobile vehicle, projection transformation is performed in which the image IMG1 is transformed into a future image IMG2 obtained at a camera viewpoint at a timing T2 that is later than the timing T1by the image delay compensation time α. A synthetic image SIMG1 is generated based on the future image IMG2 and recognition information OR on the object in the image IMG1 and is output from a display device.