Deblur and Denoise using Richardson Lucy Deconvolution for Computer Vision tasks#8038
Deblur and Denoise using Richardson Lucy Deconvolution for Computer Vision tasks#8038rzimmerdev wants to merge 22 commits intoTheAlgorithms:masterfrom
Conversation
for more information, see https://pre-commit.ci
for more information, see https://pre-commit.ci
No! What if the Kernel is empty?
Example:
>>> kernel = np.zeros((1))
>>> kernel or np.ones((3, 3))
array([[1., 1., 1.],
[1., 1., 1.],
[1., 1., 1.]])
Co-authored-by: Christian Clauss <cclauss@me.com>
algorithms-keeper
bot
added
awaiting reviews
![algorithms-keeper[bot]](https://avatars.githubusercontent.com/in/235834?s=60&v=4){kind=small}
There was a problem hiding this comment.
Click here to look at the relevant links ⬇️
🔗 Relevant Links
Repository:
Python:
Automated review generated by algorithms-keeper. If there's any problem regarding this review, please open an issue about it.
algorithms-keeper commands and options
algorithms-keeper actions can be triggered by commenting on this PR:
@algorithms-keeper reviewto trigger the checks for only added pull request files@algorithms-keeper review-allto trigger the checks for all the pull request files, including the modified files. As we cannot post review comments on lines not part of the diff, this command will post all the messages in one comment.NOTE: Commands are in beta and so this feature is restricted only to a member or owner of the organization.
| return np.sqrt(((original - reference) ** 2).mean()) | ||
|
|
||
|
|
||
| def pad_to_size(image: np.ndarray, reference: np.ndarray): |
There was a problem hiding this comment.
As there is no test file in this pull request nor any test function or class in the file computer_vision/richardson_lucy.py, please provide doctest for the function pad_to_size
Please provide return type hint for the function: pad_to_size. If the function does not return a value, please provide the type hint as: def function() -> None:
| return normalized.astype(data_type) | ||
|
|
||
|
|
||
| def gaussian_noise(size: tuple, mean=0, std=0.05): |
There was a problem hiding this comment.
Please provide return type hint for the function: gaussian_noise. If the function does not return a value, please provide the type hint as: def function() -> None:
Please provide type hint for the parameter: mean
Please provide type hint for the parameter: std
| return noise | ||
|
|
||
|
|
||
| def gaussian_filter(k: int = 5, sigma: float = 1.0) -> np.ndarray: |
There was a problem hiding this comment.
Please provide descriptive name for the parameter: k
| return estimated_img | ||
|
|
||
|
|
||
| def main(): |
There was a problem hiding this comment.
As there is no test file in this pull request nor any test function or class in the file computer_vision/richardson_lucy.py, please provide doctest for the function main
Please provide return type hint for the function: main. If the function does not return a value, please provide the type hint as: def function() -> None:
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
This comment was marked as duplicate.
This comment was marked as duplicate.
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
computer_vision/richardson_lucy.py
Outdated
| """ | ||
| estimated_img = np.full(shape=degraded.shape, fill_value=1, dtype="float64") | ||
|
|
||
| for i in range(steps): |
There was a problem hiding this comment.
| for i in range(steps): | |
| for _ in range(steps): |
i is never used in the loop body, and this is causing the pre-commit to fail
| def convolve(matrix: np.ndarray, kernel: np.ndarray) -> np.ndarray: | ||
| """ | ||
| Convolves a given kernel around a matrix through the frequency domain, | ||
| using Fourier transformations. |
There was a problem hiding this comment.
| using Fourier transformations. | |
| using Fourier transforms. |
Nitpick: "Fourier transforms", not "Fourier transformations"
| if kernel.shape[0] > matrix.shape[1] or kernel.shape[1] > matrix.shape[1]: | ||
| return matrix |
There was a problem hiding this comment.
If the kernel is too big, why not raise an error?
| input_file = str(input()).rstrip() | ||
| degraded = imageio.imread(input_file) | ||
|
|
||
| angle = int(input()) | ||
| steps = int(input()) |
There was a problem hiding this comment.
Can we get prompt messages for the input calls
|
|
||
| def pad_to_size(image: np.ndarray, reference: np.ndarray): | ||
| """Pad an image to have final shape equal to reference image.""" | ||
| p, q = (size for size in reference.shape) |
There was a problem hiding this comment.
Are these variables necessary? Why not just use reference.shape[0] and reference.shape[1]?
| return ifftshift(ifft2(np.multiply(matrix_f, kernel_f))).real | ||
|
|
||
|
|
||
| def get_motion_psf(shape: tuple, angle: float, num_pixel_dist: int = 20) -> np.ndarray: |
There was a problem hiding this comment.
| def get_motion_psf(shape: tuple, angle: float, num_pixel_dist: int = 20) -> np.ndarray: | |
| def get_motion_psf(shape: tuple[int, int], angle: float, num_pixel_dist: int = 20) -> np.ndarray: |
There was a problem hiding this comment.
Click here to look at the relevant links ⬇️
🔗 Relevant Links
Repository:
Python:
Automated review generated by algorithms-keeper. If there's any problem regarding this review, please open an issue about it.
algorithms-keeper commands and options
algorithms-keeper actions can be triggered by commenting on this PR:
@algorithms-keeper reviewto trigger the checks for only added pull request files@algorithms-keeper review-allto trigger the checks for all the pull request files, including the modified files. As we cannot post review comments on lines not part of the diff, this command will post all the messages in one comment.NOTE: Commands are in beta and so this feature is restricted only to a member or owner of the organization.
| return np.sqrt(((original - reference) ** 2).mean()) | ||
|
|
||
|
|
||
| def pad_to_size(image: np.ndarray, reference: np.ndarray): |
There was a problem hiding this comment.
As there is no test file in this pull request nor any test function or class in the file computer_vision/richardson_lucy.py, please provide doctest for the function pad_to_size
Please provide return type hint for the function: pad_to_size. If the function does not return a value, please provide the type hint as: def function() -> None:
| return normalized.astype(data_type) | ||
|
|
||
|
|
||
| def gaussian_noise(size: tuple, mean=0, std=0.05): |
There was a problem hiding this comment.
Please provide return type hint for the function: gaussian_noise. If the function does not return a value, please provide the type hint as: def function() -> None:
Please provide type hint for the parameter: mean
Please provide type hint for the parameter: std
| return noise | ||
|
|
||
|
|
||
| def gaussian_filter(k: int = 5, sigma: float = 1.0) -> np.ndarray: |
There was a problem hiding this comment.
Please provide descriptive name for the parameter: k
| return estimated_img | ||
|
|
||
|
|
||
| def main(): |
There was a problem hiding this comment.
As there is no test file in this pull request nor any test function or class in the file computer_vision/richardson_lucy.py, please provide doctest for the function main
Please provide return type hint for the function: main. If the function does not return a value, please provide the type hint as: def function() -> None:
There was a problem hiding this comment.
Click here to look at the relevant links ⬇️
🔗 Relevant Links
Repository:
Python:
Automated review generated by algorithms-keeper. If there's any problem regarding this review, please open an issue about it.
algorithms-keeper commands and options
algorithms-keeper actions can be triggered by commenting on this PR:
@algorithms-keeper reviewto trigger the checks for only added pull request files@algorithms-keeper review-allto trigger the checks for all the pull request files, including the modified files. As we cannot post review comments on lines not part of the diff, this command will post all the messages in one comment.NOTE: Commands are in beta and so this feature is restricted only to a member or owner of the organization.
| return np.sqrt(((original - reference) ** 2).mean()) | ||
|
|
||
|
|
||
| def pad_to_size(image: np.ndarray, reference: np.ndarray): |
There was a problem hiding this comment.
As there is no test file in this pull request nor any test function or class in the file computer_vision/richardson_lucy.py, please provide doctest for the function pad_to_size
Please provide return type hint for the function: pad_to_size. If the function does not return a value, please provide the type hint as: def function() -> None:
| return normalized.astype(data_type) | ||
|
|
||
|
|
||
| def gaussian_noise(size: tuple, mean=0, std=0.05): |
There was a problem hiding this comment.
Please provide return type hint for the function: gaussian_noise. If the function does not return a value, please provide the type hint as: def function() -> None:
Please provide type hint for the parameter: mean
Please provide type hint for the parameter: std
| return noise | ||
|
|
||
|
|
||
| def gaussian_filter(k: int = 5, sigma: float = 1.0) -> np.ndarray: |
There was a problem hiding this comment.
Please provide descriptive name for the parameter: k
| return estimated_img | ||
|
|
||
|
|
||
| def main(): |
There was a problem hiding this comment.
As there is no test file in this pull request nor any test function or class in the file computer_vision/richardson_lucy.py, please provide doctest for the function main
Please provide return type hint for the function: main. If the function does not return a value, please provide the type hint as: def function() -> None:
|
This PR is unfortunately taking too long and too much effort to fix, and I don't have the time because we maintainers have to clear out the backlog of PRs as much as possible before Hacktoberfest. In addition, it appears that this PR has failing doctests even when the code formatting and styling has been fixed. |
2 participants
Describe your change:
I've added the Richardson Lucy deconvolution algorithm, as well as some supporting auxiliary methods.
This algorithm is useful for recovering degraded images, given a known noise and/or blurring function.
Checklist:
Fixes: #{$ISSUE_NO}.