EAP Issue 4 Detecting Visual Errors

Skip to end of metadata
Go to start of metadata
One line summary Images with errors can only be properly identified once they have been looked at individually (some appear in Bridge to be fine, while others will show some error).Need to identify images which show pixilation, banding, other errors in the image                                                                                                                        

Detailed description Some of our images have problems with pixilation, banding and other errors. These will not always be picked up with manual checks because the thumnails in Bridge are small, we will not check each image, we are not always sure what we're looking for, etc. We need to be able to identify these across large batches of images, and ideally know what causes it. Examples:
  • EAP012 There is coloured banding across the image, and some visual information is missing (lines in drawing do not meet up at either side of the error)
  • EAP219 As above, also coloured pixels scattered throughout image
Issue champion John Salter, Matt Ruane
Possible approaches Find good (non-problematic) images and get an idea of what a 'valid' histogram looks like, then compare all images in a batch to find any outliers
Context  
AQuA Solutions  
Collections EAP;
Labels:
image image Delete
tiff tiff Delete
qa qa Delete
bit_rot bit_rot Delete
issue issue Delete
unsolved_issue unsolved_issue Delete
Enter labels to add to this page:
Please wait 
Looking for a label? Just start typing.