Adjusting Brightness/Contrast



Is there a way to overexpose images? I have cells expressing EEA1 in which CellProfiler cannot differentiate the cell boundaries, so I would like to jack up the exposure and use that image to identify cell outlines (then use the original images for analysis). Right now, I am using ImageJ to overexpose the cells (so they are completely bright and show distinctive boundaries), saving them as separate images, then uploading all images into CellProfiler.

If you have any suggestions so I can automate this process in CellProfiler, I’d greatly appreciate it!




Thanks for giving CellProfiler a try! I highly recommend that you visit our “Before Posting: Tips page” (link below). This will optimize your post so that we can best assist you.



In an image loaded into CellProfiler, you can quickly brighten any open image by right-clicking it, then choose ‘Image Contrast’ > ‘Normalize’ or ‘Log Normalize’. Note that this simply changes the values for display. If you really want to brighten the raw data (which we don’t usually suggest), then you can use the RescaleIntensity module. We don’t suggest changing the values except for display, since the thresholding algorithms are usually set to be automatically adjusted, so that even if the images are dim, but still have some contrast, then the segmentation procedure is not usually benefitted by this adjustment. There are lots of caveats to this, but in general, I suggest you just alter the images for display and let the Identify* modules do their work on the raw data.


Is there a way to adjust the contrast using a module to the level seen on the “Image contrast>Log normalise” setting?


Are you asking if you can use a module to exactly match the log rescaling output? It wouldn’t be easy to exactly match it, though you can use ImageMath to take a base-2 log of the image or raise it to an exponent to approximate the effect of log rescaling. David’s caveats above still apply though- we really only recommend this for display, not for using to segment.