Hello all,
I'm having a difficult time understanding why histogram scales change when I import .zvi using the bio-formats importer. I'd like to convert all of my files to .tif format and use ImageJ for analysis but I am concerned any subsequent analysis will not be valid as a result of this rescaling. Here's what I have done: I save a 16-bit .zvi file and load onto Mac OSX. (I know it is 16-bit because my Axiovision camera captures at this depth and the software confirms in the file information.) Using the bio-formats import option, I open the 3-channel .zvi file with the Standard-ImageJ stack option selected. The 3 channels are split; autoscaling is off, and color mode is in gray-scale. I have also tried different variations of this by using hyperstack options, and color-mode on default. The result has been the same: Upon clicking on the histogram function, I see that instead of seeing a 0-65535 range on my intensity distribution scale, images from both different channels and different samples can vary in the intensity scaling. Example: sample 1, channel 1 shows a 0-65520 scale; sample 1 channel 3 shows a 0-65520 scale ;sample 2, channel 1 shows a 0-65520 scale; sample 2, channel 3 shows a 0-58890 scale. In an attempt to confirm that the images should be on a 0-65535 scale, opening the files without bio-formats importer and then clicking on the histogram function shows me that this is in fact the case: the scale is 0-65535. (this is not an option for me however, as images areas are scrambled on the ImageJ window, i.e. the image quadrants/relative positions are 'scrambled' on the image.) It seems like a straightforward scaling or binning problem, but how can the importer be set to bin equivalently? Can anyone help me understand why this happens? How can I maintain the 16-bit scale (0-65535)? I've played around with converting to 32-bit and re-converting back to 16-bit. I'd rather avoid this step if at all necessary and retain the original data scale. Thank in advance for any help!! |
The histogram display is being scaled to the max. intensity present in the
image. Try this: open an image that gives you a scale <65,535 and display the Histogram like you have been doing, then draw one pixel of intensity 65,535 anywhere in the image using the pencil tool, then display the Histogram again. The scale should change from <65,535 to 65,535. Bio-Formats is not changing the image, so any subsequent analysis is comparable between images that have different histogram scales. -Esteban On Tue, Jul 15, 2014 at 11:33 AM, javila0624 <[hidden email]> wrote: > Hello all, > I'm having a difficult time understanding why histogram scales change when > I > import .zvi using the bio-formats importer. I'd like to convert all of my > files to .tif format and use ImageJ for analysis but I am concerned any > subsequent analysis will not be valid as a result of this rescaling. > > > > Here's what I have done: > I save a 16-bit .zvi file and load onto Mac OSX. (I know it is 16-bit > because my Axiovision camera captures at this depth and the software > confirms in the file information.) > Using the bio-formats import option, I open the 3-channel .zvi file with > the > Standard-ImageJ stack option selected. The 3 channels are split; > autoscaling > is off, and color mode is in gray-scale. I have also tried different > variations of this by using hyperstack options, and color-mode on default. > > The result has been the same: Upon clicking on the histogram function, I > see > that instead of seeing a 0-65535 range on my intensity distribution scale, > images from both different channels and different samples can vary in the > intensity scaling. Example: sample 1, channel 1 shows a 0-65520 scale; > sample 1 channel 3 shows a 0-65520 scale ;sample 2, channel 1 shows a > 0-65520 scale; sample 2, channel 3 shows a 0-58890 scale. > > In an attempt to confirm that the images should be on a 0-65535 scale, > opening the files without bio-formats importer and then clicking on the > histogram function shows me that this is in fact the case: the scale is > 0-65535. (this is not an option for me however, as images areas are > scrambled on the ImageJ window, i.e. the image quadrants/relative positions > are 'scrambled' on the image.) > > It seems like a straightforward scaling or binning problem, but how can the > importer be set to bin equivalently? > > > Can anyone help me understand why this happens? How can I maintain the > 16-bit scale (0-65535)? > > I've played around with converting to 32-bit and re-converting back to > 16-bit. I'd rather avoid this step if at all necessary and retain the > original data scale. > > > Thank in advance for any help!! > > > > > > -- > View this message in context: > http://imagej.1557.x6.nabble.com/Histogram-scale-change-when-importing-ZVI-files-tp5008753.html > Sent from the ImageJ mailing list archive at Nabble.com. > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Esteban, Thank you for the reply. A follow-up question: If my min. also varies, what does this mean? I have a few images here with min.'s of 7552, 5888 and max. of 65520. Opening without bio-formats results again in a 0-65535 distribution. The histograms reflect this with the "0-65535" image skewing more to the center, whereas the other image skews to the left. This was done using the same steps as mentioned on my post. I'd like to bin all my images equivalently to get a distribution of intensities at a given bins. This is why I'm concerned that right now, my images are not comparable. What do you think? Jorge Sent from my iPhone
|
Jorge,
You can use the "Custom Histogram" macro for your analysis. Save the macro text file at the link below. Run the macro in ImageJ with Plugins > Macro > Run, then choose the saved txt file. http://imagej.nih.gov/ij/macros/CustomHistogram.txt -Esteban On Tue, Jul 15, 2014 at 1:41 PM, javila0624 <[hidden email]> wrote: > Esteban, > Thank you for the reply. A follow-up question: > If my min. also varies, what does this mean? > > I have a few images here with min.'s of 7552, 5888 and max. of 65520. > Opening without bio-formats results again in a 0-65535 distribution. The > histograms reflect this with the "0-65535" image skewing more to the > center, whereas the other image skews to the left. > This was done using the same steps as mentioned on my post. > > I'd like to bin all my images equivalently to get a distribution of > intensities at a given bins. This is why I'm concerned that right now, my > images are not comparable. > > > What do you think? > > > Jorge > > > Sent from my iPhone > > On Jul 15, 2014, at 16:15, "G. Esteban Fernandez [via ImageJ]" < > [hidden email]<mailto: > [hidden email]>> wrote: > > The histogram display is being scaled to the max. intensity present in the > image. Try this: open an image that gives you a scale <65,535 and display > the Histogram like you have been doing, then draw one pixel of intensity > 65,535 anywhere in the image using the pencil tool, then display the > Histogram again. The scale should change from <65,535 to 65,535. > > Bio-Formats is not changing the image, so any subsequent analysis is > comparable between images that have different histogram scales. > > -Esteban > > > On Tue, Jul 15, 2014 at 11:33 AM, javila0624 <[hidden > email]</user/SendEmail.jtp?type=node&node=5008756&i=0>> wrote: > > > Hello all, > > I'm having a difficult time understanding why histogram scales change > when > > I > > import .zvi using the bio-formats importer. I'd like to convert all of my > > files to .tif format and use ImageJ for analysis but I am concerned any > > subsequent analysis will not be valid as a result of this rescaling. > > > > > > > > Here's what I have done: > > I save a 16-bit .zvi file and load onto Mac OSX. (I know it is 16-bit > > because my Axiovision camera captures at this depth and the software > > confirms in the file information.) > > Using the bio-formats import option, I open the 3-channel .zvi file with > > the > > Standard-ImageJ stack option selected. The 3 channels are split; > > autoscaling > > is off, and color mode is in gray-scale. I have also tried different > > variations of this by using hyperstack options, and color-mode on > default. > > > > The result has been the same: Upon clicking on the histogram function, I > > see > > that instead of seeing a 0-65535 range on my intensity distribution > scale, > > images from both different channels and different samples can vary in the > > intensity scaling. Example: sample 1, channel 1 shows a 0-65520 scale; > > sample 1 channel 3 shows a 0-65520 scale ;sample 2, channel 1 shows a > > 0-65520 scale; sample 2, channel 3 shows a 0-58890 scale. > > > > In an attempt to confirm that the images should be on a 0-65535 scale, > > opening the files without bio-formats importer and then clicking on the > > histogram function shows me that this is in fact the case: the scale is > > 0-65535. (this is not an option for me however, as images areas are > > scrambled on the ImageJ window, i.e. the image quadrants/relative > positions > > are 'scrambled' on the image.) > > > > It seems like a straightforward scaling or binning problem, but how can > the > > importer be set to bin equivalently? > > > > > > Can anyone help me understand why this happens? How can I maintain the > > 16-bit scale (0-65535)? > > > > I've played around with converting to 32-bit and re-converting back to > > 16-bit. I'd rather avoid this step if at all necessary and retain the > > original data scale. > > > > > > Thank in advance for any help!! > > > > > > > > > > > > -- > > View this message in context: > > > http://imagej.1557.x6.nabble.com/Histogram-scale-change-when-importing-ZVI-files-tp5008753.html > > Sent from the ImageJ mailing list archive at Nabble.com< > http://Nabble.com>. > > > > -- > > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > > > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > > > ________________________________ > If you reply to this email, your message will be added to the discussion > below: > > http://imagej.1557.x6.nabble.com/Histogram-scale-change-when-importing-ZVI-files-tp5008753p5008756.html > To unsubscribe from Histogram scale change when importing .ZVI files, > click here< > >. > NAML< > http://imagej.1557.x6.nabble.com/template/NamlServlet.jtp?macro=macro_viewer&id=instant_html%21nabble%3Aemail.naml&base=nabble.naml.namespaces.BasicNamespace-nabble.view.web.template.NabbleNamespace-nabble.view.web.template.NodeNamespace&breadcrumbs=notify_subscribers%21nabble%3Aemail.naml-instant_emails%21nabble%3Aemail.naml-send_instant_email%21nabble%3Aemail.naml > > > > > > > -- > View this message in context: > http://imagej.1557.x6.nabble.com/Histogram-scale-change-when-importing-ZVI-files-tp5008753p5008757.html > Sent from the ImageJ mailing list archive at Nabble.com. > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
Thanks a bunch, for listening and the assistance. Will follow-up with results. Jorge
|
In reply to this post by javila0624
On Jul 15, 2014, at 4:41 PM, javila0624 wrote:
> Esteban, > Thank you for the reply. A follow-up question: > If my min. also varies, what does this mean? > > I have a few images here with min.'s of 7552, 5888 and max. of 65520. Opening without bio-formats results again in a 0-65535 distribution. The histograms reflect this with the "0-65535" image skewing more to the center, whereas the other image skews to the left. > This was done using the same steps as mentioned on my post. > > I'd like to bin all my images equivalently to get a distribution of intensities at a given bins. This is why I'm concerned that right now, my images are not comparable. Run the Histogram command by pressing alt+h, enter 0 as "X min", 65536 as "X max" and you will get a histogram that uses a bin size of 256. The 0 and 65536 values are remembered so they only have to be entered the first time you press alt+h. -wayne > What do you think? > > > Jorge > > > Sent from my iPhone > > On Jul 15, 2014, at 16:15, "G. Esteban Fernandez [via ImageJ]" <[hidden email]<mailto:[hidden email]>> wrote: > > The histogram display is being scaled to the max. intensity present in the > image. Try this: open an image that gives you a scale <65,535 and display > the Histogram like you have been doing, then draw one pixel of intensity > 65,535 anywhere in the image using the pencil tool, then display the > Histogram again. The scale should change from <65,535 to 65,535. > > Bio-Formats is not changing the image, so any subsequent analysis is > comparable between images that have different histogram scales. > > -Esteban > > > On Tue, Jul 15, 2014 at 11:33 AM, javila0624 <[hidden email]</user/SendEmail.jtp?type=node&node=5008756&i=0>> wrote: > >> Hello all, >> I'm having a difficult time understanding why histogram scales change when >> I >> import .zvi using the bio-formats importer. I'd like to convert all of my >> files to .tif format and use ImageJ for analysis but I am concerned any >> subsequent analysis will not be valid as a result of this rescaling. >> >> >> >> Here's what I have done: >> I save a 16-bit .zvi file and load onto Mac OSX. (I know it is 16-bit >> because my Axiovision camera captures at this depth and the software >> confirms in the file information.) >> Using the bio-formats import option, I open the 3-channel .zvi file with >> the >> Standard-ImageJ stack option selected. The 3 channels are split; >> autoscaling >> is off, and color mode is in gray-scale. I have also tried different >> variations of this by using hyperstack options, and color-mode on default. >> >> The result has been the same: Upon clicking on the histogram function, I >> see >> that instead of seeing a 0-65535 range on my intensity distribution scale, >> images from both different channels and different samples can vary in the >> intensity scaling. Example: sample 1, channel 1 shows a 0-65520 scale; >> sample 1 channel 3 shows a 0-65520 scale ;sample 2, channel 1 shows a >> 0-65520 scale; sample 2, channel 3 shows a 0-58890 scale. >> >> In an attempt to confirm that the images should be on a 0-65535 scale, >> opening the files without bio-formats importer and then clicking on the >> histogram function shows me that this is in fact the case: the scale is >> 0-65535. (this is not an option for me however, as images areas are >> scrambled on the ImageJ window, i.e. the image quadrants/relative positions >> are 'scrambled' on the image.) >> >> It seems like a straightforward scaling or binning problem, but how can the >> importer be set to bin equivalently? >> >> >> Can anyone help me understand why this happens? How can I maintain the >> 16-bit scale (0-65535)? >> >> I've played around with converting to 32-bit and re-converting back to >> 16-bit. I'd rather avoid this step if at all necessary and retain the >> original data scale. >> >> >> Thank in advance for any help!! >> >> >> >> >> >> -- >> View this message in context: >> http://imagej.1557.x6.nabble.com/Histogram-scale-change-when-importing-ZVI-files-tp5008753.html >> Sent from the ImageJ mailing list archive at Nabble.com<http://Nabble.com>. >> >> -- >> ImageJ mailing list: http://imagej.nih.gov/ij/list.html >> > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html > > > ________________________________ > If you reply to this email, your message will be added to the discussion below: > http://imagej.1557.x6.nabble.com/Histogram-scale-change-when-importing-ZVI-files-tp5008753p5008756.html > To unsubscribe from Histogram scale change when importing .ZVI files, click here< > NAML<http://imagej.1557.x6.nabble.com/template/NamlServlet.jtp?macro=macro_viewer&id=instant_html%21nabble%3Aemail.naml&base=nabble.naml.namespaces.BasicNamespace-nabble.view.web.template.NabbleNamespace-nabble.view.web.template.NodeNamespace&breadcrumbs=notify_subscribers%21nabble%3Aemail.naml-instant_emails%21nabble%3Aemail.naml-send_instant_email%21nabble%3Aemail.naml> > > > > > -- > View this message in context: http://imagej.1557.x6.nabble.com/Histogram-scale-change-when-importing-ZVI-files-tp5008753p5008757.html > Sent from the ImageJ mailing list archive at Nabble.com. > > -- > ImageJ mailing list: http://imagej.nih.gov/ij/list.html -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html |
This works perfectly! Thank you Wayne. Thank you Esteban. Great suggestions, and you've cleared part of my confusion.
|
Free forum by Nabble | Edit this page |