{
  "id": 55451,
  "title": "3-band Geotiff images show up blank (dark)!",
  "url": "/competitions/dstl-satellite-imagery-feature-detection/discussion/55451",
  "author_name": "",
  "post_date": "2018-04-26T17:46:28.736000300Z",
  "votes": -1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I am having a  problem with 48 bit depth 3-band geotiff image file with an extension *.tif. The image file opens fine with displaying full range of RGB but shows up completely blank (or black, if you will) when opened with other regular image viewers such as 'Paint' or 'Photos' etc.\nWhen I inspected the pixel values in matlab (through imread), the intensity values had very narrrow dynamic range per channel between 250 to 1000! It explained why the image shows dark , but the confusing part is the same image shows perfectly in QGIS. I need to perform classification on these images but I am not sure if it would work with such a narrow dynamic range. Again the frustrating part is that the images shows perfectly vibrant with full range of RGB colourmap in QGIS.\nPlease help me where I am going wrong reading the 3-band images.\nThank you.</p>",
  "messages": [
    {
      "id": "319722",
      "postDate": "04/26/2018 17:46:28",
      "content": "<p>I am having a  problem with 48 bit depth 3-band geotiff image file with an extension *.tif. The image file opens fine with displaying full range of RGB but shows up completely blank (or black, if you will) when opened with other regular image viewers such as 'Paint' or 'Photos' etc.\nWhen I inspected the pixel values in matlab (through imread), the intensity values had very narrrow dynamic range per channel between 250 to 1000! It explained why the image shows dark , but the confusing part is the same image shows perfectly in QGIS. I need to perform classification on these images but I am not sure if it would work with such a narrow dynamic range. Again the frustrating part is that the images shows perfectly vibrant with full range of RGB colourmap in QGIS.\nPlease help me where I am going wrong reading the 3-band images.\nThank you.</p>",
      "rawMarkdown": "I am having a  problem with 48 bit depth 3-band geotiff image file with an extension *.tif. The image file opens fine with displaying full range of RGB but shows up completely blank (or black, if you will) when opened with other regular image viewers such as 'Paint' or 'Photos' etc.\nWhen I inspected the pixel values in matlab (through imread), the intensity values had very narrrow dynamic range per channel between 250 to 1000! It explained why the image shows dark , but the confusing part is the same image shows perfectly in QGIS. I need to perform classification on these images but I am not sure if it would work with such a narrow dynamic range. Again the frustrating part is that the images shows perfectly vibrant with full range of RGB colourmap in QGIS.\nPlease help me where I am going wrong reading the 3-band images.\nThank you.",
      "votes": null
    },
    {
      "id": "327261",
      "postDate": "05/11/2018 06:40:11",
      "content": "<p>You might already figured this out, but QGIS automatically enhance the contrast of the images for displaying purpose, while other viewers that you used simply loads the unprocessed values. In QGIS, you can right click, then select 'Properties' on the raster layer, you can see that the 'Contrast Enhancement' is 'Stretched to MinMax'. If you select 'No enhancement', you should get back the same \"black image\".</p>\n\n<p>Note that contrast enhancement is usually for visualization purpose only. </p>",
      "rawMarkdown": "You might already figured this out, but QGIS automatically enhance the contrast of the images for displaying purpose, while other viewers that you used simply loads the unprocessed values. In QGIS, you can right click, then select 'Properties' on the raster layer, you can see that the 'Contrast Enhancement' is 'Stretched to MinMax'. If you select 'No enhancement', you should get back the same \"black image\".\n\nNote that contrast enhancement is usually for visualization purpose only.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 327261,
      "author_name": "alwinkai",
      "author_url": "",
      "post_date": "05/11/2018 06:40:11",
      "content": "<p>You might already figured this out, but QGIS automatically enhance the contrast of the images for displaying purpose, while other viewers that you used simply loads the unprocessed values. In QGIS, you can right click, then select 'Properties' on the raster layer, you can see that the 'Contrast Enhancement' is 'Stretched to MinMax'. If you select 'No enhancement', you should get back the same \"black image\".</p>\n\n<p>Note that contrast enhancement is usually for visualization purpose only. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "319722": "I am having a  problem with 48 bit depth 3-band geotiff image file with an extension *.tif. The image file opens fine with displaying full range of RGB but shows up completely blank (or black, if you will) when opened with other regular image viewers such as 'Paint' or 'Photos' etc.\nWhen I inspected the pixel values in matlab (through imread), the intensity values had very narrrow dynamic range per channel between 250 to 1000! It explained why the image shows dark , but the confusing part is the same image shows perfectly in QGIS. I need to perform classification on these images but I am not sure if it would work with such a narrow dynamic range. Again the frustrating part is that the images shows perfectly vibrant with full range of RGB colourmap in QGIS.\nPlease help me where I am going wrong reading the 3-band images.\nThank you.",
    "327261": "You might already figured this out, but QGIS automatically enhance the contrast of the images for displaying purpose, while other viewers that you used simply loads the unprocessed values. In QGIS, you can right click, then select 'Properties' on the raster layer, you can see that the 'Contrast Enhancement' is 'Stretched to MinMax'. If you select 'No enhancement', you should get back the same \"black image\".\n\nNote that contrast enhancement is usually for visualization purpose only."
  },
  "source": "meta"
}