Wide-Range Image Generation via Evaluation-Based Subject Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image capture devices struggle to generate wide-range images beyond their hardware limitations, requiring users to manually rotate the camera and combine multiple images, which can result in inconsistent subject appearances and reduced image quality.
Innovation Solution
An image processing device that acquires images while moving in a predetermined direction, detects same subject images, calculates evaluation values, decides on the best subject images for combination based on these values, and generates a wide-range image by combining them, ensuring a more attractive and coherent final image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of moving object
If multiple images are captured and combined to create a wide-range image, then the image capture angle is improved, but the subject appearance consistency deteriorates
Solution Approach 1:
The patent changes the selection parameter from simple chronological ordering to evaluation-based selection. By calculating evaluation values for each captured image based on subject detection results, the system selects only images with high evaluation values (indicating good subject appearance) for combination, thus maintaining subject consistency while achieving wide-range coverage.
Solution Approach 2:
The system automatically evaluates and selects the best images for combination without requiring manual user input. The evaluation unit autonomously calculates evaluation values based on subject detection, and the combination unit automatically selects images with high evaluation values, making the system self-sufficient in maintaining image quality.
2Area of moving object
If multiple images are captured and combined to create a wide-range image, then the image capture angle is improved, but the image quality deteriorates
Solution Approach 1:
The patent introduces evaluation values as a new parameter to assess image quality. By calculating evaluation values that reflect subject appearance quality and using this parameter to guide image selection, the system ensures that only high-quality images are combined, preventing quality deterioration despite capturing multiple images.
Solution Approach 2:
The system uses evaluation values as feedback to guide the image combination process. The evaluation unit provides quality assessment feedback to the combination unit, which then selects images with high evaluation values, creating a feedback loop that maintains image quality throughout the wide-range image generation process.
3Ease of operation
If characteristic points are used to combine images, then the image combination process is simplified, but the subject appearance consistency deteriorates
Solution Approach 1:
The patent segments the image combination process into two independent stages: (1) subject detection and evaluation, and (2) image combination based on evaluation values. This segmentation allows the system to use simple characteristic point matching for alignment while separately ensuring subject appearance consistency through evaluation-based selection, combining simplicity with quality.
Data Source
AI summary
An image acquisition unit 52 acquires images consecutively captured while an image capture range is moved in a predetermined direction. A face detection unit 53 detects same subject images from the images acquired. A facial expression determination unit 55 calculates evaluation values of each of the same subject images detected. A facial expression decision unit 56 decides, as a combination target, a specific subject image from the same subject images, based on the evaluation values calculated. A combination unit 58 generates a wide-range image by combining the subject image decided as the combination target with the images sequentially acquired by the image acquisition unit 52.