Video Summary Device Edge Processing Dash Cam Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Processing and reviewing extensive video data captured by dash cams in vehicles is time-consuming and resource-intensive, consuming network, storage, and computing resources.
Innovation Solution
A video summary device determines the relevance of different portions of time in video data based on events detected, generates a summary by selecting frames from the most relevant time ranges, and provides this summary to a user device, thereby preserving resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the user device processes and reviews all video data captured by dash cams, then complete video review is achieved, but time consumption and resource usage (network, storage, computing) increase significantly
Solution Approach 1:
The patent segments the video review process into two distinct stages: (1) edge device performs preliminary processing to generate condensed video summaries with detected objects and events, and (2) user device reviews only these condensed summaries. This segmentation allows complete video review to be achieved at the edge while user review time is dramatically reduced.
Solution Approach 2:
The edge device performs preliminary video processing actions before the user needs to review the content. By pre-processing video data to extract key information (objects, events, timestamps) and generating condensed summaries, the system prepares the review material in advance, eliminating the need for the user device to process raw video data during review.
2Reliability
If the user device processes and reviews all video data captured by dash cams, then complete video review is achieved, but computing resources are consumed excessively
Solution Approach 1:
The patent segments the computing workload by function and device capability. The edge device (dashboard system) handles computationally intensive tasks including video frame processing, object detection, event identification, and summary generation. The user device only handles lightweight display and interaction tasks for reviewing the pre-processed summaries, dramatically reducing its computing resource consumption.
Solution Approach 2:
The condensed video summary acts as an intermediary representation between the raw video data and the user review process. Instead of the user device directly processing raw video, it interacts with the structured summary data (containing object detections, event timestamps, and key frame references), which requires minimal computing resources to render and review.
3Reliability
If extensive video data is transmitted to the user device, then complete video information is available, but network resources are consumed significantly
Solution Approach 1:
The system extracts only the essential information from the extensive video data before transmission. The edge device extracts object detections, event timestamps, location data, and key frame references, condensing hours of video into structured summary data. This extracted information is then transmitted to the user device, maintaining information availability while minimizing network bandwidth consumption.
Solution Approach 2:
The patent changes the parameter representation of video information from raw pixels and frames to structured metadata (JSON format containing object classifications, bounding boxes, timestamps, and event types). This parameter transformation reduces data volume by orders of magnitude while preserving the essential information needed for review, thereby reducing network transmission energy consumption.
Data Source
AI summary
A video summary device may determine, based on video data, a plurality of events that occurred during a period of time. The video summary device may determine first measures of relevance of a plurality of portions of the period of time. A first measure of relevance of a portion of time may be determined based on one or more events of the plurality of events. The video summary device may determine second measures of relevance of a plurality of ranges of time between different portions of time of the plurality of portions of the period of time. The second measures of relevance is determined based on the first measures of relevance determined for the different portions of time. The video summary device may determine particular frames of the video data, as a video summary of the video data, based on the second measures of relevance.