Motion-Based Image Stitching Using Gyroscopic Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing panoramic photography systems face challenges such as determining appropriate exposure settings, blurring, parallax issues, aligning images, and correcting perspective changes, especially in handheld devices where computational costs for image analysis and registration become prohibitive as the number of images increases.
Innovation Solution
A method that aligns digital images using motion data from gyroscopic and accelerometer sensors, eliminating the need for image analysis and registration, by determining relative motion and generating a perspective transform matrix based on image capture parameters and motion data to align images in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard image analysis and registration operations are performed to align panoramic images, then alignment accuracy is improved, but computational cost becomes prohibitive as the number of images increases
Solution Approach 1:
The patent applies preliminary action by using motion data from gyroscopic and accelerometer sensors to pre-determine the relative motion between images before alignment is needed. This motion data is integrated to obtain instantaneous position information that predicts how images should be aligned, eliminating the need for computationally expensive post-capture image analysis and registration operations.
Solution Approach 2:
The patent replaces the mechanical/image-processing system with a sensor-based system. Instead of using image analysis algorithms to determine alignment, the system substitutes this with motion sensors (gyroscopes and accelerometers) that directly measure the physical motion of the camera device, converting an image processing problem into a motion measurement problem.
2Area of stationary object
If the number of images captured within a given time period is increased to improve panoramic coverage, then field of view is improved, but computational cost of image processing becomes prohibitive
Solution Approach 1:
The patent extracts the alignment computation from the image processing pipeline and relocates it to the sensor data processing domain. By separating the motion measurement function (performed by inertial sensors) from the image alignment function, the system can handle multiple images without proportionally increasing computational cost, as the motion data provides direct alignment information without requiring analysis of each image.
3Productivity
If image capture rate is increased to enable real-time panoramic photography, then productivity is improved, but complexity of aligning images in real-time increases
Solution Approach 1:
The patent implements self-service by using the motion sensors to automatically provide alignment information without requiring complex external processing. The gyroscopic and accelerometer sensors continuously track device motion and self-generate the transformation data needed for image alignment, making the system self-sufficient and reducing the computational burden regardless of capture rate.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for efficient and low-computational-cost image stitching, enabling high image capture rates and reducing the complexity of aligning panoramic images in real-time, particularly in resource-constrained devices like mobile phones and tablets.
Implementation Method 1
motion information for the first image may be obtained (e.g., from a gyroscopic and/or accelerometer sensors)
Implementation Method 2
motion information for the first image may be obtained (e.g., from a gyroscopic and/or accelerometer sensors)
Data Source
AI summary
Systems, methods, and computer readable media for stitching or aligning multiple images (or portions of images) to generate a panoramic image are described. In general, techniques are disclosed for using motion data (captured at substantially the same time as image data) to align images rather than performing image analysis and/or registration operations. More particularly, motion data may be used to identify the rotational change between successive images. The identified rotational change, in turn, may be used to generate a transform that, when applied to an image allows it to be aligned with a previously captured image. In this way, images may be aligned in real-time using only motion data.