# Automated Rectification of Image
Implements the modified version of the following paper:
[Chaudhury, Krishnendu, Stephen DiVerdi, and Sergey Ioffe. "Auto-rectification
of user photos." 2014 IEEE International Conference on Image Processing (ICIP).
IEEE, 2014.](https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/42532.pdf)
Modifcation note: Instead of finding edge direction using structural tensor and its eigenvectors as in paper, I have used more reliable canny edge detection and probabalistic hough line transform.
## Results
Input image:
After rectification:
## How it works
First, compute list of 'edgelets'. An edgelet is a tuple of edge location, edge direction and edge strength.
```python
edgelets1 = compute_edgelets(image)
vis_edgelets(image, edgelets1) # Visualize the edgelets
```
Next, find dominant vanishing point using ransac algorithm. In our case it turns out to be horizontal.
```python
vp1 = ransac_vanishing_point(edgelets1, num_ransac_iter=2000,
threshold_inlier=5)
vp1 = reestimate_model(vp1, edgelets1, threshold_reestimate=5)
vis_model(image, vp1) # Visualize the vanishing point model
```
Remove the inliers for horizontal vanishing point. Vertical lines should now be dominant. Recompute the vanishing point using ransac should give us vertical vanishing point.
```python
edgelets2 = remove_inliers(vp1, edgelets1, 10)
vp2 = ransac_vanishing_point(edgelets2, num_ransac_iter=2000,
threshold_inlier=5)
vp2 = reestimate_model(vp2, edgelets2, threshold_reestimate=5)
vis_model(image, vp2) # Visualize the vanishing point model
```
Finally, compute homography and warp the image so that we have a fronto parellel view with orthogonal axes:
```
warped_img = compute_homography_and_warp(image, vp1, vp2,
clip_factor=clip_factor)
```
Raw data
{
"_id": null,
"home_page": "https://github.com/SajjadAemmi/PyVP",
"name": "pyvp",
"maintainer": "",
"docs_url": null,
"requires_python": "",
"maintainer_email": "",
"keywords": "",
"author": "Sajjad Aemmi",
"author_email": "sajjadaemmi@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/b3/3a/270cf6fa05f1fc79fdcfb319efe809d33fb8beb588e354e5e29d89a6b0dc/pyvp-0.1.1.tar.gz",
"platform": null,
"description": "# Automated Rectification of Image\n\nImplements the modified version of the following paper: \n\n[Chaudhury, Krishnendu, Stephen DiVerdi, and Sergey Ioffe. \"Auto-rectification\nof user photos.\" 2014 IEEE International Conference on Image Processing (ICIP).\n IEEE, 2014.](https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/42532.pdf)\n\nModifcation note: Instead of finding edge direction using structural tensor and its eigenvectors as in paper, I have used more reliable canny edge detection and probabalistic hough line transform.\n\n## Results\n\nInput image:\n\n\n\nAfter rectification:\n\n\n\n## How it works\n\nFirst, compute list of 'edgelets'. An edgelet is a tuple of edge location, edge direction and edge strength. \n\n```python\nedgelets1 = compute_edgelets(image)\nvis_edgelets(image, edgelets1) # Visualize the edgelets\n```\n\n\n\nNext, find dominant vanishing point using ransac algorithm. In our case it turns out to be horizontal.\n\n```python\nvp1 = ransac_vanishing_point(edgelets1, num_ransac_iter=2000, \n threshold_inlier=5)\nvp1 = reestimate_model(vp1, edgelets1, threshold_reestimate=5)\nvis_model(image, vp1) # Visualize the vanishing point model\n```\n\n\n\nRemove the inliers for horizontal vanishing point. Vertical lines should now be dominant. Recompute the vanishing point using ransac should give us vertical vanishing point. \n\n```python\nedgelets2 = remove_inliers(vp1, edgelets1, 10)\nvp2 = ransac_vanishing_point(edgelets2, num_ransac_iter=2000,\n threshold_inlier=5)\nvp2 = reestimate_model(vp2, edgelets2, threshold_reestimate=5)\nvis_model(image, vp2) # Visualize the vanishing point model\n```\n\n\n\nFinally, compute homography and warp the image so that we have a fronto parellel view with orthogonal axes: \n\n```\nwarped_img = compute_homography_and_warp(image, vp1, vp2,\n clip_factor=clip_factor)\n```\n\n\n",
"bugtrack_url": null,
"license": "",
"summary": "Vanishing Point Detector",
"version": "0.1.1",
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "d321faf41716685e709b9af721a8df3f8cc233dbc1d44e2d606047387d5e3d01",
"md5": "1bf052da02e375604ad0bcfde43f887b",
"sha256": "c55856e7aa4a8077a305bebdf47f3ae2edacad5d28c44f29f6e25a729e818fdc"
},
"downloads": -1,
"filename": "pyvp-0.1.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "1bf052da02e375604ad0bcfde43f887b",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": null,
"size": 7588,
"upload_time": "2023-03-28T09:15:22",
"upload_time_iso_8601": "2023-03-28T09:15:22.519760Z",
"url": "https://files.pythonhosted.org/packages/d3/21/faf41716685e709b9af721a8df3f8cc233dbc1d44e2d606047387d5e3d01/pyvp-0.1.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "b33a270cf6fa05f1fc79fdcfb319efe809d33fb8beb588e354e5e29d89a6b0dc",
"md5": "ae9b33c125f0a9a81be77e052f219ebd",
"sha256": "41c922edc6b8afd9f4f87ba69034cd708f5c20c4a2f32942804d713d4844a7d3"
},
"downloads": -1,
"filename": "pyvp-0.1.1.tar.gz",
"has_sig": false,
"md5_digest": "ae9b33c125f0a9a81be77e052f219ebd",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 7485,
"upload_time": "2023-03-28T09:15:24",
"upload_time_iso_8601": "2023-03-28T09:15:24.620560Z",
"url": "https://files.pythonhosted.org/packages/b3/3a/270cf6fa05f1fc79fdcfb319efe809d33fb8beb588e354e5e29d89a6b0dc/pyvp-0.1.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-03-28 09:15:24",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "SajjadAemmi",
"github_project": "PyVP",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [],
"lcname": "pyvp"
}
JSON
