{
  "id": 69740,
  "title": "Seam carving to reduce image size - alternative to resampling?",
  "url": "/competitions/human-protein-atlas-image-classification/discussion/69740",
  "author_name": "Vig",
  "post_date": "2018-10-26T16:46:51.881000",
  "votes": 3,
  "comment_count": 0,
  "views": 0,
  "content": "<p>The most challenging part of this competition is the sheer size of the images (512x512) and the fact that the features cover only a couple of pixels. On the other hand, there are lot of voids in these images. So the question is, can we get rid of the voids and reduce the image size. </p>\n\n<p>In <a href=\"https://www.kaggle.com/vignam/seam-carving-to-reduce-image-size\">my kernel</a>, I show how for some train set example this may work. But, with a better idea, I hope it can be used for all the images. I think this would be far better way of reducing the image size than resampling due to the fine structure of the features.</p>",
  "messages": [
    {
      "id": 410782,
      "postDate": "2018-10-26T16:46:51.880Z",
      "content": "<p>The most challenging part of this competition is the sheer size of the images (512x512) and the fact that the features cover only a couple of pixels. On the other hand, there are lot of voids in these images. So the question is, can we get rid of the voids and reduce the image size. </p>\n\n<p>In <a href=\"https://www.kaggle.com/vignam/seam-carving-to-reduce-image-size\">my kernel</a>, I show how for some train set example this may work. But, with a better idea, I hope it can be used for all the images. I think this would be far better way of reducing the image size than resampling due to the fine structure of the features.</p>",
      "rawMarkdown": "The most challenging part of this competition is the sheer size of the images (512x512) and the fact that the features cover only a couple of pixels. On the other hand, there are lot of voids in these images. So the question is, can we get rid of the voids and reduce the image size. \n\nIn [my kernel][1], I show how for some train set example this may work. But, with a better idea, I hope it can be used for all the images. I think this would be far better way of reducing the image size than resampling due to the fine structure of the features.\n\n\n  [1]: https://www.kaggle.com/vignam/seam-carving-to-reduce-image-size",
      "votes": 3
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "410782": "The most challenging part of this competition is the sheer size of the images (512x512) and the fact that the features cover only a couple of pixels. On the other hand, there are lot of voids in these images. So the question is, can we get rid of the voids and reduce the image size. \n\nIn [my kernel][1], I show how for some train set example this may work. But, with a better idea, I hope it can be used for all the images. I think this would be far better way of reducing the image size than resampling due to the fine structure of the features.\n\n\n  [1]: https://www.kaggle.com/vignam/seam-carving-to-reduce-image-size"
  }
}