Image Editing Apparatus Extracting Key Frames From Movement Locus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-speed continuously shot images of moving objects result in a heavy load for analysis processing and are less effective unless they include images shot at the object's highest or lowest position, necessitating a mechanism to appropriately extract key images from a large volume of data.
Innovation Solution
An image editing apparatus comprising an image storage unit, moving position detection unit, locus-of-movement detection unit, feature image specifying unit, and image extraction unit to identify and extract images corresponding to feature points in the locus of movement, allowing for effective learning material generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If high-speed continuous shooting is used to capture moving objects, then the quantity of images increases, but the processing load increases and analysis becomes less effective
Solution Approach 1:
The patent extracts only the essential images from continuous shooting sequences by detecting feature points in the movement locus and identifying images that contain these feature points. This extraction process removes unnecessary images while preserving the critical information needed for analysis, thereby reducing processing load without sacrificing analytical effectiveness.
Solution Approach 2:
The patent applies different processing strategies to different portions of the image sequence. Instead of uniformly processing all images, it identifies specific regions of interest (feature points in movement locus) and selectively extracts images based on their relevance to these local features, optimizing the balance between quantity and processing efficiency.
2Quantity of substance
If all continuously shot images are used for learning, then the volume of learning materials increases, but the effectiveness decreases unless key positions are captured
Solution Approach 1:
The system extracts images that specifically contain feature points representing key positions in the movement locus (such as highest or lowest positions). By taking out only these critical images rather than using all continuous shots, the system maintains high effectiveness for learning purposes while reducing the overall volume of materials to manageable levels.
Solution Approach 2:
The patent changes the selection parameter from uniform sampling of all images to selective sampling based on feature point detection. By using the movement locus information as a selection criterion, the system transforms the image selection process to prioritize quality and relevance over quantity, ensuring that captured images represent key moments in the object's motion.
Data Source
AI summary
An image editing method includes detecting positions of the moving object contained in the images for the stored series of images, detecting a locus of movement of the moving object based on the detected positions of the moving object contained in the series of images, specifying an image corresponding to a feature point of the detected locus of movement of the moving object, extracting, from the series of images stored in the memory, images which include the specified image and correspond to a second shooting time interval longer than the first shooting time interval, and storing the extracted series of images as an image file in a memory.