TDI Image Sensor Offset Correction via Cross-Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing TDI image sensors face limitations in integrating over a large number of lines due to imperfections such as instability, micro-vibrations, optical distortions, and alignment errors, which lead to blurring of the final image, and require non-standard, precisely aligned sensors, making them difficult to produce using standard components.
Innovation Solution
The solution involves intrinsic shake detection directly from image data using cross-correlation calculations between vectors from multiple TDI detection matrices, allowing for correction of offsets and integration without the need for additional sensors, enabling production from standard, commercially available components and relaxing alignment requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If integration is performed over a large number of lines to improve signal-to-noise ratio, then image quality improves, but cumulative imperfections from carrier instability, micro-vibrations, optical distortions, and alignment errors cause unacceptable blurring
Solution Approach 1:
The patent divides the integration process into multiple segments, each performed over a limited number of lines (e.g., N lines per detector matrix). By segmenting the total integration into multiple smaller integration stages across several matrices, the cumulative alignment errors are reset between segments, preventing the accumulation of blurring while maintaining high signal-to-noise ratio through the combined effect of multiple segments.
Solution Approach 2:
The patent introduces an optical motion sensor (shake detector) as an intermediary device that measures carrier movements between successive image acquisitions. This intermediary sensor provides measurement data that is used to correct alignment errors in the final image composite, thereby compensating for the blurring effects without limiting the integration line count.
2Duration of action of moving object
If multiple TDI detection matrices are used to extend integration range, then integration over more lines is enabled, but near-perfect alignment between sensors and image detection arrays is required, making monolithic co-integration necessary
Solution Approach 1:
The patent introduces an optical motion sensor as an intermediary measurement device that captures carrier movement data. This intermediary sensor enables the use of multiple separately-aligned TDI matrices by providing the correction data needed to compensate for alignment imperfections, thereby extending integration time without requiring monolithic co-integration.
Solution Approach 2:
The patent implements a feedback mechanism where the optical motion sensor continuously measures carrier movements, and this measurement information is fed back to correct the image data from multiple TDI matrices. The feedback loop allows the system to compensate for alignment errors dynamically, enabling the use of multiple matrices with relaxed alignment tolerances.
3Ease of manufacture
If standard commercially available components are used to simplify manufacturing, then ease of manufacture improves, but existing solutions require non-standard motion detectors and precise alignment
Solution Approach 1:
The patent makes the TDI detector matrices multi-functional by using them both for their primary image detection function and for measuring carrier movements through cross-correlation of their output signals. This eliminates the need for separate non-standard motion detectors, allowing the use of standard TDI matrices while achieving both ease of manufacture and extended integration capability.
Solution Approach 2:
The patent enables the TDI detector matrices to self-measure carrier movements by performing cross-correlation on their own output signals. This self-service capability eliminates the need for external specialized motion sensors, allowing the system to be built from standard components while maintaining the ability to correct alignment errors.
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 enables integration over a larger number of lines, correcting for optical distortions and alignment errors, while simplifying the manufacturing process by using standard components and reducing the need for precise alignment, resulting in improved image quality and cost-effectiveness.
Implementation Method 1
calculate cross correlations between image data vectors provided by the different image detection matrices; deduce a shift between said image data vectors
Implementation Method 2
each image detection matrix comprises a plurality of parallel photosensitive pixel lines... configured to provide at its output image data vectors obtained by accumulation of image data generated by said pixel lines
Data Source
Figure 1
Figure 2A~3
Figure 2B
AI summary
An image sensor (IMS) intended to be carried on a platform (SAT) moving above a scene to be observed, said sensor comprising a plurality of TDI-type image detection matrices (M1, M2), aligned along the direction of movement (D) of said platform, and a signal processing circuit (SPC) for the signals generated by said matrices, in which said signal processing circuit is configured to: - calculate cross-correlations between image data vectors (VD1, VD2) provided by the different image detection matrices; - deduce an offset between said image data vectors; - correct said offset by re-regulating said image vectors; and - integrate the re-registered image data vectors to provide at its output an accumulated image data vector. A spaceborne or airborne optical image acquisition instrument comprising such an image sensor.