Image Analysis TRIZ Cases
- Neural Network Chip De-Convolution Hardware Circuit
A neural network chip performs expansion and zero-padding on feature maps within its data reading circuit.
- Object Detection Using Pixel-Luminance Differences
A detection device uses weak discriminators to evaluate luminance differences between two pixels for rapid object identification.
- AR Labeling via Aligned 3D CAD Models
A system aligns virtual objects with physical items using augmented reality to generate precise pixel-level annotations.
- Track Identification via Bolt Head Orientation
A system identifies undercarriage tracks by analyzing bolt head orientation data to generate unique joint signatures.
- Image Search Heat Map Generation for Subject Localization
An image processing apparatus extracts feature vectors and maps to generate a heat map indicating search target likelihood.
- Stereo Imaging Disparity Map Edge Refinement
A stereo imaging system refines disparity map edge coordinates using dedicated detection and refinement modules.
- In-Vehicle Line-of-Sight Calibration Device with Dynamic Eye Detection
A calibration device adjusts line-of-sight detection processing using dynamic eye detectability checks to select between face-moving and stationary calibration modes.
- 3D Mapping Using Single-Sensor Speckle Pattern Analysis
A single image sensor captures primary speckle patterns projected by a coherent light source to derive accurate 3D object maps.
- Augmented Reality Wireless Signal Simulation
An augmented reality system simulates wireless signal coverage using three-dimensional virtual representations of the environment.
- Polyline Construction from Line Segments for Localization
Merging neighboring line segments into continuous polylines reduces processing time while improving localization accuracy.
- Mixed Reality Virtual Object Rendering for Real-World Collision Avoidance
Information processing apparatus generates virtual objects to notify users of real-world obstacles in mixed reality environments.
- 6DoF Media Guide Voice Generation via Viewpoint Adaptation
A media distribution device generates guide voices describing rendered images from user viewpoints in virtual spaces.
- Autonomous Image Annotation via Knowledge Base Decomposition
A machine learning system decomposes images into parts and stores relationships in a knowledge base to enable autonomous image annotation.
- Contour Extraction Apparatus for Image Processing
A contour extracting unit isolates area objects from images using user-defined selection conditions received via an operating unit.
- White Spot Mura Detection in Display Panels
Multi-class support vector machine classifies image patches to detect white spot Mura defects, resolving low accuracy in high-resolution display inspections.
- Merchandise Return Support System Using Image Recognition
A return support server acquires merchandise images to identify item state and determine return eligibility.
- 3D Shape Measurement Using Segmented Line Patterns
A 3D shape measurement apparatus projects segmented line patterns to rapidly acquire and correlate image data for efficient spatial reconstruction.
- Vehicle Image Processing Resolving 3D Object Misrecognition
A vehicle-mounted image processing device distinguishes road surface features from three-dimensional objects using positional relationships.
- Geolocation Estimation via Ground-to-Overhead Image Transformation
A system transforms ground-level images into overhead views for location estimation.