Seeker Image Object Detection via Optics Pivoting and PSF Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for controlling missiles struggle to quickly and accurately recognize objects with weak optical signatures in noisy environments, as they require extensive scanning time and are prone to image smearing due to camera optics movement.
Innovation Solution
The method involves pivoting the camera optics during image integration and using image processing to recalculate the search image with the point spread function and pivoting movement, allowing for efficient compensation of image smearing and improved object recognition, even in noisy backgrounds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the camera optics are stationary during image integration, then image sharpness is maintained, but scanning time increases significantly
Solution Approach 1:
The patent applies the Dynamics principle by making the camera optics movable during image integration. Instead of keeping the optics stationary, the system pivots the optics continuously throughout the integration period, allowing multiple scene regions to be captured within a single exposure. This dynamic approach reduces the number of separate images needed for scanning, thereby reducing total scanning time while the computational processing compensates for the motion-induced blur.
2Productivity
If the camera optics are pivoted during image integration, then scanning time is reduced, but image smearing occurs
Solution Approach 1:
The patent implements the Feedback principle by using an inertial measurement unit (IMU) to continuously track the actual pivoting motion of the optics during image integration. This motion data is fed back to the image processing unit, which uses it to compute compensation functions. The feedback loop allows the system to correct for motion-induced smearing by adjusting the pixel coordinate transformations based on the actual observed pivoting trajectory, thereby maintaining image quality while enabling fast scanning.
Solution Approach 2:
The patent replaces the mechanical approach of keeping optics stationary with a computational approach. Instead of using a stationary optical system that would require multiple sequential exposures, the system uses a moving optical system combined with computational image processing. The mechanical pivoting motion is substituted with an equivalent computational transformation that reconstructs the image data, allowing fast scanning without permanent image degradation.
3Reliability
If a large number of images are captured for comprehensive scanning, then detection accuracy improves, but processing time increases
Solution Approach 1:
The patent applies the Continuity of useful action principle by capturing image data continuously during the pivoting motion without stopping to take separate stationary images. The optics pivot continuously throughout the integration period, and the detector continuously records photons across different angular positions. This continuous capture approach consolidates what would otherwise require multiple discrete image acquisitions into a single uninterrupted exposure, reducing processing time while maintaining comprehensive scene coverage for accurate detection.
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 significantly reduces scanning time, enhances image sharpness, and enables better detection of weakly signed objects by accounting for optics movement and noise, allowing for accurate object recognition and marking for missile guidance.
Implementation Method 1
a detector having a plurality of detector cells and readout electronics for reading out the detector cells, wherein an image of the environment is integrated during a defined integration period from a start of integration to an end of integration on the detector
Data Source
Figure 1~3
AI summary
The invention relates to a method for detecting an object (4) in a search image (20), in which the object (4) and an environment around the object (4) are recorded by an optic (8) of a camera (6). It is proposed that the optic (8) is pivoted during the integration of a recording image (28) in the environment and that an image processing unit (14) uses the dot spreading function and the pivoting movement of the optic (8) to reconstruct the search image (20) from the recording image (28) and searches the search image (20) for the object (4).