Video Frame Selector for Item Identification in Dynamic Scenes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current mobile devices struggle to effectively identify and provide information about items in dynamic video frames, requiring users to hold devices steady or record videos instead of taking pictures, which can result in blurry images.

Innovation Solution

A method and system that uses a video processor application with a video frame selector module, item identification module, and location-based incentive module to identify items in video frames, tag them, and offer related offers or incentives based on geographic location, utilizing image recognition algorithms and GPS data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If users take pictures with mobile devices, then they can capture items, but the user has to hold the mobile device steady and the object needs to remain static to avoid blurry pictures

Engineering Contradiction:
Improvepicture clarityVSAvoiduser operation requirement
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The system transitions from static image capture to dynamic video frame processing. Instead of requiring a steady hand and static object for a clear picture, the system processes multiple video frames to identify items, allowing natural hand movement and object motion while maintaining identification accuracy through frame selection and image recognition algorithms.

Inventive Principle:
Principle #15Dynamics

2Adaptability or versatility

If users record video instead of taking pictures, then they can capture dynamic scenes, but the resulting video frames may be blurry

Engineering Contradiction:
Improvedynamic scene captureVSAvoidimage clarity
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system segments the video into individual frames and selectively processes only certain frames for item identification. By analyzing motion between frames and selecting frames with minimal motion blur, the system captures dynamic scenes effectively while maintaining image clarity for accurate item recognition.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary analysis of video frames to identify those with minimal motion blur before proceeding with item identification. By pre-selecting optimal frames based on motion criteria, the system ensures that subsequent image recognition operates on high-quality images, resolving the clarity issue in dynamic video capture.

Inventive Principle:
Principle #10Preliminary action

3Loss of information

If the system processes every video frame, then it can identify all items, but it increases processing time and computational resources

Engineering Contradiction:
Improveitem identification completenessVSAvoidprocessing time
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The system processes only a subset of video frames rather than every frame. By selecting frames based on motion criteria and item presence indicators, the system achieves sufficient item identification completeness while significantly reducing processing time and computational resource requirements.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system divides the video processing task into frame selection and item identification stages. By first filtering frames based on motion and relevance criteria, then applying image recognition only to selected frames, the system maintains comprehensive item identification while minimizing overall processing time through selective computation.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3131255B1Video processing system for identifying items in video frames
Publication Date: 2019.10.02 EBAY INC
  • EP3131255B1 patent drawingFigure 1
  • EP3131255B1 patent drawingFigure 2~3
  • EP3131255B1 patent drawingFigure 4

AI summary

A method and a system are provided. The system comprises a processor-implemented video frame selector module configured to determine a video frame to process that is received from a mobile device, the processor-implemented video frame selector module comprising a video frame analyzer module configured to determine a difference between a first video frame and a second video frame, and a video frame tag module configured to tag the first or second video frame as the determined video frame for item identification when the difference exceeds a predetermined amount. The system further comprises a processor-implemented item identification module configured to identify an item in the determined video frame and to tag the determined video frame with an identification of the item. The system also comprises a processor-implemented market module configured to generate offers of the identified item from at least one merchant to the mobile device.