Circular scanning technique for large area inspection
Described embodiments provide a method of generating an image of a region of interest of a target object. A plurality of concentric circular scan trajectories are determined to sample the region of interest. Each of the concentric circular scan trajectories have a radius incremented from an innermost concentric circular scan trajectory having a minimum radius to an outermost concentric circular scan trajectory having a maximum radius. A number of samples are determined for each of the concentric circular scan trajectories. A location of each sample is determined for each of the concentric circular scan trajectories. The locations of each sample are substantially uniformly distributed in a Cartesian coordinate system of the target object. The target object is iteratively rotated along each of the concentric circular scan trajectories and images are captured at the determined sample locations to generate a reconstructed image from the captured images.
1. A method of generating an image of a region of interest of a target object by an imaging system, the method comprising:
determining a plurality of concentric circular scan trajectories to sample the region of interest, each of the plurality of concentric circular scan trajectories having a radius incremented by a pitch value from an innermost concentric circular scan trajectory having a minimum radius to an outermost concentric circular scan trajectory having a maximum radius;
determining a number of samples for each of the plurality of concentric circular scan trajectories;
wherein the determining a number of samples for each of the plurality of concentric circular scan trajectories comprises:
determining, for each concentric circular scan trajectory, an angle increment and a radius increment; and determining, based upon the determined angle increment and the determined radius increment, a number of samples, a rotation speed, and a plurality of rotation angles for each concentric circular scan trajectory;
determining a location of each sample for each of the plurality of concentric circular scan trajectories, the locations of each sample substantially uniformly distributed in a Cartesian coordinate system of the target object to reduce image distortion;
iteratively rotating the target object along each of the concentric circular scan trajectories and capturing images at the determined sample locations; and
generating a reconstructed image from the captured images.
2. The method of claim 1 , wherein rotating the target object comprises:
rotating the target object at a determined constant angular velocity, the determined constant angular velocity reducing vibration of the target object.
3. The method of claim 1 , wherein rotating the target object comprises:
rotating the target object at a determined constant linear velocity.
4. The method of claim 1 , wherein the region of interest is circular, and the maximum radius is substantially equal to a radius of the region of interest.
5. The method of claim 1 , wherein determining a location of each sample for each of the plurality of concentric circular scan trajectories comprises:
mapping each sample location to Cartesian coordinates; and
interpolating one or more neighboring sample locations.
6. The method of claim 5 , wherein the interpolating comprises one of:
nearest-neighbor interpolation; or
linear interpolation.
7. The method of claim 1 , further comprising:
performing a simulated annealing search to optimize the one or more concentric circular scan trajectories.
8. The method of claim 1 , wherein each of the plurality of rotation angles for each concentric circular scan trajectory is associated with a sample location.
9. The method of claim 1 , further comprising:
constraining at least one of: angular motion, rotational motion and pixel coverage area to interpolate one or more neighboring sample locations to overlap pixels on neighboring concentric circular scan trajectories.
10. The method of claim 1 , wherein generating a reconstructed image from the captured images comprises:
performing super resolution (SR) on one or more of the captured images to generate a high resolution output image wherein performing super resolution comprises:
capturing a sequence of low resolution images for each concentric circular scan trajectory;
performing iterative backpropagation to generate one or more super resolution images having sub-pixel resolution of corresponding ones of the sequence of low resolution images; and
transforming the one or more super resolution images from a polar coordinate system to a Cartesian coordinate system.
11. The method of claim 10 , further comprising:
performing mosaicing of the one or more transformed super resolution images to generate a high resolution wide field of view composite output image.
12. The method of claim 11 , wherein performing mosaicing comprises:
stitching together one or more super resolution images for each concentric circular scan trajectory.
13. The method of claim 12 , comprising:
stitching together one or more super resolution images for each concentric circular scan trajectory independently of other concentric circular scan trajectories.
14. The method of claim 10 , further comprising:
reducing blurring and noise effects in the sequence of low resolution images by performing truncating singular value decomposition.
15. The method of claim 10 , wherein capturing the sequence of low resolution images comprises:
dividing each concentric circular scan trajectory into segments, each segment having a determined radial resolution and a determined angular resolution; and
applying a regular shift in sub-pixel steps in a radial direction for each concentric circular scan trajectory to acquire low resolution images.
16. The method of claim 15 , wherein the regular shift step is based upon a pixel size of the high resolution output image.
17. The method of claim 1 , further comprising:
synchronizing a camera frame rate of the imaging system, an illumination level of the imaging system, a translational movement speed of a target stage of the imaging system and a rotational movement speed of the target stage.
18. An imaging system for generating an image of a region of interest of a target object, the imaging system comprising:
a camera configured to capture images of the target object; an illumination source configured to illuminate the target object; a target stage configured to receive the target object, the target stage configured to provide a translational movement and a rotational movement of the target object; and a controller configured to:
determine a plurality of concentric circular scan trajectories to sample the region of interest, each of the plurality of concentric circular scan trajectories having a radius incremented from an innermost concentric circular scan trajectory having a minimum radius to an outermost concentric circular scan trajectory having a maximum radius;
determine a number of samples for each of the plurality of concentric circular scan trajectories;
wherein the controller is configured to:
determine, for each concentric circular scan trajectory, an angle increment and a radius increment; and
determine, based upon the determined angle increment and the determined radius increment, a number of samples, a rotation speed, and a plurality of rotation angles for each concentric circular scan trajectory;
determine a location of each sample for each of the plurality of concentric circular scan trajectories, the locations of each sample substantially uniformly distributed in a Cartesian coordinate system of the target object to reduce image distortion;
control the camera and target stage to iteratively rotate the target object along each of the concentric circular scan trajectories and capture images at the determined sample locations; and
generate a reconstructed image from the captured images.
19. The imaging system of claim 18 , wherein the target stage is configured to rotate the target object at one of:
a determined constant angular velocity, the determined constant angular velocity reducing vibration of the target object; or
a determined constant linear velocity.
20. The imaging system of claim 18 , wherein the region of interest is circular, and the maximum radius is substantially equal to a radius of the region of interest.
21. The imaging system of claim 18 , wherein the controller is configured to:
map each sample location to Cartesian coordinates; and
interpolate one or more neighboring sample locations.
22. The imaging system of claim 21 , wherein the controller is configured to interpolate one or more neighboring sample locations by one of:
nearest-neighbor interpolation; or
by linear interpolation.
23. The imaging system of claim 18 , wherein the controller is configured to:
perform a simulated annealing search to optimize the one or more concentric circular scan trajectories.
24. The imaging system of claim 18 , wherein each of the plurality of rotation angles for each concentric circular scan trajectory is associated with a sample location.
25. The imaging system of claim 18 , wherein the controller is configured to:
constrain at least one of: angular motion, rotational motion and pixel coverage area to interpolate one or more neighboring sample locations to overlap pixels on neighboring concentric circular scan trajectories.
26. The imaging system of claim 18 , wherein the controller is configured to:
perform super resolution (SR) on one or more of the captured images to generate a high resolution output image.
27. The imaging system of claim 26 , wherein the controller is configured to:
capture a sequence of low resolution images for each concentric circular scan trajectory;
perform iterative backpropagation to generate one or more super resolution images having sub-pixel resolution of corresponding ones of the sequence of low resolution images; and
transform the one or more super resolution images from a polar coordinate system to a Cartesian coordinate system.