Harrison Wood

Thanks for the help, Nick. Perhaps one last question- My template constructions have generally been failing on their first iteration, even when specifying an initial template with the -z option. Failed builds look like this: Unaligned images, just overlaid. If I try the template construction again without changing the output prefix, then it succeeds. This works even if the input files are different. I suspect there are some files left over from the failed run which somehow help out with avoiding...

Template building is inherently registration-based and if corresponding structures are not contained in all the images, then you're going to have artifacts, like the ones you mention, where those correspondences don't exist. Makes sense! For my purposes, though, are the warps themselves likely to be inaccurate? I feel like the actual coordinate warps are correct- even if the resulting template image is poor. For establishing a common coordinate system to compare relative locations, the result is...

Template building is inherently registration-based and if corresponding structures are not contained in all the images, then you're going to have artifacts, like the ones you mention, where those correspondences don't exist. Makes sense! For my purposes, though, are the warps themselves likely to be inaccurate? I feel like the actual coordinate warps are correct- even if the resulting template image is poor. For establishing a common coordinate system to compare relative locations, the result is...

Template building is inherently registration-based and if corresponding structures are not contained in all the images, then you're going to have artifacts, like the ones you mention, where those correspondences don't exist. Makes sense! For my purposes, though, are the warps themselves likely to be inaccurate? I feel like the actual coordinate warps are correct- even if the resulting template image is poor. For establishing a common coordinate system to compare relative locations, the result is...

Hi Nick, You can see an example of the sort of brain image data I'm looking to make at http://engertlab.fas.harvard.edu:4001/#/home. I've uploaded an explanation of the issues I'm having (with plenty of examples!) with the template construction script to https://www.scribd.com/document/364922474/ANTS-Problems. It seems to be working mostly well so far, but there are a few things worth smoothing out I feel. The problems I linked earlier are a bit of a read, so it's OK if you have more important things...

Hi Nick, I've run the template construction with some success, but I'm getting weird boundaries. Any suggestions for, say, blurring out these boundaries and correcting the brightness of the non-mutually-imaged areas to have the same intensity as the universally captured areas? I've got multiple scans of slightly different regions in different fish, and would like to build a big map. They're particularly sharp at the edges of acquisitions. Though I know the script is primarily for aligning multiple...

Hi Nick, I've run the template construction with some success, but I'm getting weird boundaries. Any suggestions for, say, blurring out these boundaries and correcting the brightness of the non-mutually-imaged areas to have the same intensity as the universally captured areas? I've got multiple scans of slightly different regions in different fish, and would like to build a big map. They're particularly sharp at the edges of acquisitions. Though I know the script is primarily for aligning multiple...

Hi Nick, I've run the template construction with some success, but I'm getting weird boundaries. Any suggestions for, say, blurring out these boundaries and correcting the brightness of the non-mutually-imaged areas to have the same intensity as the universally captured areas? I've got multiple scans of slightly different regions in different fish, and would like to build a big map. They're particularly sharp at the edges of acquisitions. Though I know the script is primarily for aligning multiple...