Panoramic Image Generation Using Motion Vector Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional panoramic image photography is complex and requires manual proficiency, often resulting in image distortion due to the need for separate devices and post-processing, and struggles to accurately align overlapping images captured while rotating a camera.
Innovation Solution
A method and apparatus that estimate camera motion to detect and select regions for panoramic image generation without additional hardware, using a Motion Vector detector and region selector to align and blend images in real-time, minimizing distortion and enabling immediate image review.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a camera is manually mounted to an auxiliary apparatus such as a tripod and rotated for photographing, then panoramic images can be captured, but the operation becomes complex and requires user proficiency
Solution Approach 1:
The camera system performs panoramic photographing automatically without requiring manual mounting to auxiliary apparatus. The camera self-manages the panoramic capture process through automated control, eliminating the need for users to manually operate tripods or rotation mechanisms.
Solution Approach 2:
The patent replaces the mechanical system of manual tripod mounting and rotation with an automated electronic control system. The camera uses internal mechanisms and software control to achieve panoramic capture, substituting the complex mechanical operation with a simplified electronic automation process.
2Area of stationary object
If images are captured by rotating a camera in horizontal or vertical direction, then panoramic coverage is achieved, but perspective distortion occurs making overlapping portions difficult to estimate
Solution Approach 1:
The camera performs preliminary estimation of camera motion (pan and tilt angles) before capturing images. This preliminary action allows the system to pre-calculate the overlapping regions between consecutive images, ensuring accurate alignment without perspective distortion issues.
Solution Approach 2:
The system uses feedback from the estimated camera motion to adjust the image capture and processing. By continuously monitoring the pan and tilt angles and using this information to guide the stitching process, the system compensates for perspective distortion and achieves accurate alignment of overlapping portions.
3Adaptability or versatility
If conventional panoramic processing is performed with separate post-processing, then images can be combined, but the process requires separate devices and additional time
Solution Approach 1:
The patent merges the panoramic image processing function into the camera itself. The camera integrates both the capture and processing capabilities in a single device, combining what were previously separate operations (capture with tripod/rotation and post-processing with external equipment) into one unified system.
Solution Approach 2:
The camera is designed with multi-functionality, serving both as the capture device and the processing device. The same camera that captures the individual images also performs the panoramic stitching and processing, eliminating the need for separate specialized equipment.
4Loss of time
If real-time panoramic image generation is implemented, then immediate image review is enabled, but processing speed and accuracy requirements increase
Solution Approach 1:
The system performs preliminary estimation of camera motion between frames in real-time as images are captured. This preliminary action enables immediate processing without waiting for complete capture sequences, reducing post-processing time while maintaining accuracy through continuous motion estimation.
Data Source
AI summary
A method and apparatus for photographing a panoramic image. The method includes detecting a Motion Vector (MV) of a plurality of images captured through a photographing unit, selecting regions to be included in the panoramic image from the plurality of images, based on the MV, and generating the panoramic image including the selected regions.


