Summary Image Browsing via Object Segmentation and 3D Trajectory Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multimedia browsing systems are inefficient in managing and rapidly finding desired multimedia content, such as images or videos, due to the vast amount of data they need to handle, making it difficult for users to conveniently recognize search results.
Innovation Solution
A method and system for generating and aligning summary images based on object segments extracted from videos, where each segment is synthesized with a background image along its motion trajectory, allowing for display in a 3D structure with varying sizes and colors, enabling users to easily recognize object movements and interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multimedia browsing systems manage vast amounts of multimedia data, then the quantity of data handled increases, but user convenience in recognizing search results deteriorates
Solution Approach 1:
The patent segments multimedia content by extracting individual objects from video frames and creating separate object segments. Each object is tracked independently through the video sequence, allowing the system to manage vast amounts of data by breaking it down into discrete, searchable object instances rather than treating the entire video as a single unit.
Solution Approach 2:
The patent adds a temporal dimension to object representation by tracking objects across multiple video frames. Objects are represented not just as static images but as sequences of segments with motion trajectories, creating a time-aware visualization that helps users recognize patterns and interactions in large multimedia datasets.
2Measurement precision
If objects are extracted and displayed with varying sizes and colors along motion trajectories, then object recognition improves, but system complexity increases
Solution Approach 1:
The patent applies local quality by assigning different visual properties (colors, sizes) to different parts of the visualization based on their specific characteristics. Each object segment's color and size are determined by local attributes such as object category, confidence level, or temporal position, allowing precise object recognition through differentiated visual encoding without requiring complex global processing.
Data Source
AI summary
An embodiment of the present invention provides a summary image browsing system and method. A summary image browsing method of the present invention may comprise the steps of: tracking motion trajectory of an object from an input video; extracting the object from the input video and then generating a series of object segments; and synthesizing the series of object segments with a background image along the motion trajectory of the object and generating a summary image having a thickness according to an occurrence time interval for each object extracted from the input video.