Image Shake Correction Using Motion Vector Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image shake correction technologies face challenges in accurately determining the angular velocity of objects during panning, leading to residual image shake due to offset errors from angular velocity sensors, which can be caused by drift, temperature changes, and individual differences.
Innovation Solution
An image shake correction device that acquires angular velocity information from multiple detectors and motion vectors to calculate the object's angular velocity with high precision, using a control unit to adjust the image blur correction, thereby reducing shake residuals by integrating lens and camera angular velocity data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If angular velocity sensor output is used directly for shake correction, then the correction can be performed, but offset errors from drift and temperature changes cause shake residual
Solution Approach 1:
The patent uses feedback by detecting the actual motion vector from the image and comparing it with the angular velocity sensor output. The difference (offset error) is then fed back to correct the sensor output, continuously improving measurement precision while maintaining reliable shake correction.
Solution Approach 2:
The patent replaces the purely mechanical/angular velocity sensor-based system with an optical-based correction system. By calculating motion vectors from image analysis, the system substitutes the sensor's mechanical measurement with an optical measurement, eliminating sensor drift and offset errors.
2Ease of operation
If shift lens is moved to correct speed difference, then panning assist is achieved, but large offset component causes shake residual
Solution Approach 1:
The patent extracts the offset error component from the angular velocity sensor output by comparing it with the motion vector derived from image analysis. By separating and removing this erroneous offset component, the system maintains panning assist functionality while eliminating shake residual.
Solution Approach 2:
The motion vector calculation from image analysis serves as an intermediary that mediates between the angular velocity sensor output and the actual object motion. This intermediary allows the system to correct sensor errors while maintaining the desired panning assist functionality.
3Measurement precision
If multiple angular velocity sensors are used, then measurement precision improves, but device complexity increases
Solution Approach 1:
The patent introduces image analysis as an intermediary system that provides an alternative measurement of object motion. This intermediary allows the system to cross-validate sensor readings and correct errors without requiring multiple physical sensors, thus improving precision while controlling complexity.
Data Source
AI summary
In an imaging apparatus, an interchangeable lens can be mounted on a body. The interchangeable lens includes a lens angular velocity sensor and acquires angular velocity information regarding shake. A camera body includes a camera angular velocity sensor in which an offset is smaller than in the lens angular velocity sensor to acquire angular velocity information. A motion vector detection unit detects a motion vector based on a plurality of frame images. An object angular velocity calculation unit performs a process of extracting an object using angular velocity information and a motion vector by the camera angular velocity sensor and calculates an angular velocity of the object by the imaging apparatus from the angular velocity information by the lens angular velocity sensor. A camera control unit and a lens control unit perform image blur correction control of the object during exposure using the calculated angular velocity of the object.