Terminal Device Image Analysis Before Cropping for Gesture Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile phone terminal devices struggle to recognize images independently of the camera's zoom state, leading to inappropriate processing modes when using the camera unit for tasks like barcode recognition or video games.
Innovation Solution
A terminal device acquires image data, performs analysis prior to cropping, detects specified images or hand gestures, and executes corresponding processing, allowing for control of image capture and display regardless of the zoom state, enabling instructions through hand signs or motions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image analysis is performed after cropping the image data for display, then the processing is simpler and faster, but the specified image (e.g., hand gesture) may not be detected when it is outside the cropped display area
Solution Approach 1:
The patent performs image analysis on the entire captured image data before cropping it for display. This preliminary analysis ensures that hand gestures or specified images located anywhere in the full capture area are detected, even if they fall outside the final cropped display region. The system analyzes the complete image frame first, then selectively crops and displays only the relevant portion, thereby maintaining detection accuracy without requiring analysis of the entire high-resolution image.
Solution Approach 2:
The patent segments the image processing into distinct stages: first analyzing the entire captured image to detect hand gestures or specified images, then separately cropping the image data for display purposes. This segmentation allows the detection function to operate on the full image area while the display function operates on a reduced subset, resolving the contradiction between comprehensive detection and simplified processing.
2Measurement precision
If the camera unit zooms in to display a detailed view of a subject (e.g., face), then the display shows high detail, but hand gestures or signs for control may be outside the displayed area and undetectable
Solution Approach 1:
The system performs hand gesture detection on the entire captured image frame before cropping it for detailed display. This allows users to make hand gestures anywhere within the full camera view, including areas that will be cropped out of the final detailed display. The detection occurs preliminarily on the complete image, ensuring that control gestures remain detectable even when the display shows only a zoomed-in portion of the scene.
Solution Approach 2:
The patent separates the detection dimension from the display dimension. The hand gesture detection operates on the full-dimensional captured image frame, while the display presents a cropped subset. This dimensional separation allows the system to maintain full gesture control flexibility in the detection space while providing detailed views in the display space, effectively resolving the contradiction between zoom detail and gesture accessibility.
3Measurement precision
If the entire captured image is analyzed for hand gestures, then gesture detection accuracy is maintained, but more processing power and time are required
Solution Approach 1:
The patent performs image analysis on the entire captured image frame before cropping, but only to the extent necessary for detection purposes. The system analyzes the full image area for hand gestures or specified images, then proceeds to crop the image data for display. This preliminary analysis approach maintains detection accuracy across the entire field of view while limiting the processing scope to detection-only operations on the full frame, rather than performing complex processing on the entire high-resolution image data.
Data Source
AI summary
An information processing apparatus that acquires image data captured by an image capturing device; performs an analysis on the image data prior to cropping the image data for display to detect a specified image from the image data; generates an image for display by cropping the image data; controls a display to display the image for display; and executes predetermined processing when the specified image is detected in the image data as a result of performing the analysis on the image data prior to the cropping.