Here is a discussion I found that explains what .tif format are and how to deal with them.
\nQuestion: should we use .tif format or convert them to .jpg knowing of the risk of deterioration of the quality of the image.
\nFeel free to give me your feedback it would be very helpful.
Here is a discussion I found that explains what .tif format are and how to deal with them.
\nQuestion: should we use .tif format or convert them to .jpg knowing of the risk of deterioration of the quality of the image.
\nFeel free to give me your feedback it would be very helpful.
Great pointer!
\nI just posted a quick notebook to open WSI images using the OpenSlide library. cuCIM library is another option for directly loading WSI images.
", "rawMarkdown": "Great pointer!\n\nI just posted a [quick notebook](https://www.kaggle.com/code/naotous/mayo-eda-whole-slide-images/notebook) to open WSI images using the OpenSlide library. cuCIM library is another option for directly loading WSI images.", "votes": 3, "replies": [ { "id": 1846203, "postDate": "2022-07-06T23:20:48.967Z", "content": "Saw your notebook very helpful thank you for sharing.
", "rawMarkdown": "Saw your notebook very helpful thank you for sharing.", "votes": 1 } ] }, { "id": 1850862, "postDate": "2022-07-10T20:42:20.767Z", "content": "I would say this thread takes it very far: https://www.kaggle.com/competitions/mayo-clinic-strip-ai/discussion/335976
\nalso, you may check my conversion kernel: https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-eda-load-crop-images
You can use this docker image to read the images
\ndocker pull vikash112/pathology
I will try it out.
", "rawMarkdown": "I will try it out. " } ] }, { "id": 1846231, "postDate": "2022-07-07T01:10:03.427Z", "content": "Thank you for tread, in my opinion convertation to jpeg is useless, if we can read tiff format in general this is everything some tensors of numbers and mostly we are reduce image size to 512x512, so if we lost some information from convertation it not be so important after reduce to 512x512
\nUpd but may be to reduce memory 🤔
", "rawMarkdown": "Thank you for tread, in my opinion convertation to jpeg is useless, if we can read tiff format in general this is everything some tensors of numbers and mostly we are reduce image size to 512x512, so if we lost some information from convertation it not be so important after reduce to 512x512\n\nUpd but may be to reduce memory 🤔", "votes": 1, "replies": [ { "id": 1846233, "postDate": "2022-07-07T01:12:54.760Z", "content": "yeah you're right I will try both approaches just in case.
", "rawMarkdown": "yeah you're right I will try both approaches just in case.", "votes": 1 }, { "id": 1846234, "postDate": "2022-07-07T01:14:44.320Z", "content": "Give me know if I wrong
", "rawMarkdown": "Give me know if I wrong", "votes": 1 } ] } ], "comments": [ { "id": 1846201, "author_name": "Naoto Usuyama", "author_url": "", "post_date": "2022-07-06T23:17:24.823000", "content": "Great pointer!
\nI just posted a quick notebook to open WSI images using the OpenSlide library. cuCIM library is another option for directly loading WSI images.
", "votes": 3, "replies": [ { "id": 1846203, "author_name": "Med Ali Bouchhioua", "author_url": "", "post_date": "2022-07-06T23:20:48.967000", "content": "Saw your notebook very helpful thank you for sharing.
", "votes": 1, "replies": [] } ] }, { "id": 1850862, "author_name": "Jirka", "author_url": "", "post_date": "2022-07-10T20:42:20.767000", "content": "I would say this thread takes it very far: https://www.kaggle.com/competitions/mayo-clinic-strip-ai/discussion/335976
\nalso, you may check my conversion kernel: https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-eda-load-crop-images
You can use this docker image to read the images
\ndocker pull vikash112/pathology
I will try it out.
", "votes": 0, "replies": [] } ] }, { "id": 1846231, "author_name": "Kartushov Danil", "author_url": "", "post_date": "2022-07-07T01:10:03.427000", "content": "Thank you for tread, in my opinion convertation to jpeg is useless, if we can read tiff format in general this is everything some tensors of numbers and mostly we are reduce image size to 512x512, so if we lost some information from convertation it not be so important after reduce to 512x512
\nUpd but may be to reduce memory 🤔
", "votes": 1, "replies": [ { "id": 1846233, "author_name": "Med Ali Bouchhioua", "author_url": "", "post_date": "2022-07-07T01:12:54.760000", "content": "yeah you're right I will try both approaches just in case.
", "votes": 1, "replies": [] }, { "id": 1846234, "author_name": "Kartushov Danil", "author_url": "", "post_date": "2022-07-07T01:14:44.320000", "content": "Give me know if I wrong
", "votes": 1, "replies": [] } ] } ], "raw_markdown_by_id": { "1846165": "Here is a [discussion](https://www.kaggle.com/competitions/geolifeclef-2022-lifeclef-2022-fgvc9/discussion/311983) I found that explains what .tif format are and how to deal with them.\n\nQuestion: should we use .tif format or convert them to .jpg knowing of the risk of deterioration of the quality of the image.\nFeel free to give me your feedback it would be very helpful.", "1846201": "Great pointer!\n\nI just posted a [quick notebook](https://www.kaggle.com/code/naotous/mayo-eda-whole-slide-images/notebook) to open WSI images using the OpenSlide library. cuCIM library is another option for directly loading WSI images.", "1850862": "I would say this thread takes it very far: https://www.kaggle.com/competitions/mayo-clinic-strip-ai/discussion/335976\nalso, you may check my conversion kernel: https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-eda-load-crop-images", "1847225": "You can use this docker image to read the images \ndocker pull vikash112/pathology", "1846231": "Thank you for tread, in my opinion convertation to jpeg is useless, if we can read tiff format in general this is everything some tensors of numbers and mostly we are reduce image size to 512x512, so if we lost some information from convertation it not be so important after reduce to 512x512\n\nUpd but may be to reduce memory 🤔" } }