Electric vehicle data based video composition and content augmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of sorting through hours of video recordings from electric vehicle cameras to identify relevant video fragments for creating a composite video journal is time-consuming and computationally demanding, and deleting older video files to make room for new ones can result in the loss of valuable data.
Innovation Solution
Utilizing electric vehicle data and artificial intelligence modeling to identify and protect the most valuable video fragments, generate a composite video, and insert context-suitable augmented content, while managing storage capacity to prioritize retention of important footage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual sorting of video recordings is performed to identify relevant fragments, then user control over video selection is maintained, but time consumption and computational resources increase significantly
Solution Approach 1:
The system automatically processes video recordings by analyzing EV data and generating composite videos without requiring manual user intervention. The processor autonomously identifies relevant video fragments, determines scene types, selects appropriate fragments, and assembles composite videos, allowing the system to serve itself rather than requiring continuous user operation.
Solution Approach 2:
The patent replaces manual mechanical sorting operations with automated electronic processing. Instead of users manually reviewing and selecting video fragments, the system uses processors to automatically analyze video data, determine scene types through AI models, and select relevant fragments based on EV contextual data, substituting human effort with computational automation.
2Reliability
If all video recordings are retained for potential use, then data loss is prevented, but storage capacity is quickly exhausted
Solution Approach 1:
The system extracts only the most relevant and valuable video fragments from the complete video recordings. By analyzing EV data such as acceleration, braking, location changes, and user interactions, the system identifies and extracts specific meaningful moments, storing only these extracted fragments rather than the entire video files, thus reducing storage requirements while preserving important data.
Solution Approach 2:
The patent applies different retention priorities to different portions of video data based on their significance. Instead of uniformly retaining or deleting all video segments, the system evaluates each fragment's importance using EV contextual data and scene type analysis, preserving high-value fragments (such as those containing notable events or user interactions) while allowing less important segments to be deleted or not stored.
3Productivity
If AI modeling is used to automatically identify video fragments, then processing time is reduced, but computational resources and system complexity increase
Solution Approach 1:
The patent segments the video processing task into distinct functional modules: EV data collection, video fragment identification, scene type determination using AI models, fragment selection, and composite video generation. Each module handles a specific aspect of the processing pipeline, allowing for optimized computation at each stage and reducing the complexity burden on any single component while maintaining high overall productivity.
4Measurement precision
If video fragments are selected based on EV data and scene type analysis, then relevant content is identified accurately, but processing complexity and computational load increase
Solution Approach 1:
The patent introduces EV contextual data as an intermediary layer between raw video recordings and the selection process. Instead of directly analyzing video content for relevance, the system uses EV data (acceleration, braking, location, user interactions) as a mediator to identify and flag potentially relevant time segments, which are then subjected to scene type analysis. This intermediary approach improves selection accuracy while managing computational complexity by filtering video data through multiple progressive stages.
Data Source
AI summary
Technical solutions present systems and methods for generating a composite video of a trip and inserting augmented content using AI modeling and vehicle data. A solution can identify a plurality of videos taken from a vehicle between a first time and a second time. The solution can identify, for the plurality of videos, a plurality of video fragments, each one of which corresponding to data of the vehicle at a time interval of a plurality of time intervals between the first time and the second time. The solution can determine, based on the plurality of video fragments input into a model, a type of scene for each video fragment and select, a set of video fragments based on the respective data and the respective type of scene of a plurality of sets of video fragments to generate a composite video using the set of video fragments.