FAQ: What should I do before start using ImageJ in my project?
This section is dedicated to new users just starting with imaging in the hope that it will help save time and resources.
There are quite a few things that you can do.
- Install ImageJ, read the documentation available, the tutorials in the web and play a bit with the program. Have a look at the Built-in Macro Functions and get used to the Macro Recorder (go to Plugins>Macros>Record…) to learn how to create macros while using the GUI and how to call the different functions of the program.
- Update to the latest version. The installations downloaded from the ImageJ site do not necessarily contain the latest bug fixes. Please read this FAQ (How do I update ImageJ).
- One can save a lot of time by planning a project before even starting to collect images. Image capture is sometimes very time consuming compared to some of the automated analyses ImageJ can do, so making sure that you got the right type of image is a great start.
While image processing is very powerful and can do wonderful things, there is always a limit of what can be achieved. Sometimes it is not possible to convert poor quality images into good quality, reliable results (for instance object segmentation commonly depends on image type and quality). It is advisable to spend some time testing and organising how image capture is going to proceed before embarking into a large scale data collection.
- Do yourself a great favour and do not use lossy-compressed images. JPEG compression is good to store digital pictures for the Web because they are small, but it is not good at all to do serious imaging. JPEG compression throws away information that the eye does not see, but ImageJ will, so there is a good chance of ending up measuring artifacts instead of data (especially when analysing textures and pixel intensities). This situation is, however, avoidable: while most digital cameras save in JPEG format by default, it is likely that they can also save images in some non-lossy format (such as TIFF or a custom RAW format). Use those formats instead, you can convert them later to TIFF (compressed or uncompressed) or PNG (which was designed to achieve non-lossy compression). Be aware that once an image has been saved as compressed JPEG there is no way of reverting to the original (so an old JPEG-compressed image saved again as TIFF still contains all the JPEG compression artifacts of the original).
If you do not believe that the artifacts are there, take an RGB JPEG-compressed image, convert it to HSB and then take a look at the blockiness especially of the Hue channel .
From left to right: Jpeg compressed RGB image (original 100×155 pixels), the Hue channel, the Saturation channel and the Brightness channel.
If your papers get rejected because the referees don't like image analysis on JPEGs, don't complain!
- Think in advance what image resolution is required to analyse the objects of interest and whether the microscope and camera resolutions can provide that. While magnification is how much bigger than the original the sample appears under the microscope, resolution is the ability to distinguish between two points on an image (giving the amount of detail that can be observed). Many modern digital cameras used for microscopy can digitise more detail than the microscope objective resolution (this is given by the objective's numerical aperture), so “high resolution” digital images may not contain the level of detail one thinks they do (this is called empty magnification). Dealing with unnecessarily large images will affect processing time and memory requirements. Some plugins may not run on very large images. So, if you are doing microscopy, find out the resolution of each of your objectives to avoid processing images with empty magnification.
- If you are doing bright-field microscopy, make sure that you do background illumination correction while acquiring. A method to correct uneven background illumination is described in detail in this FAQ). Note that after acquisition, this type of correction cannot be easily applied because it is based on the illumination settings of your microscope at the time of image capture.
- There is a friendly mailing list where one can ask questions, but it may save some of your time to first search the mail archive. It is very likely that your question or a similar one may have been answered in the past.
- Look at the collection of plugins and macros already available. There is a large number of plugins, many of these are really excellent. Whatever you want to do, it may have been done before, so there is a chance that a plugin or macro already exists.
Good luck with your project. When you come up with some interesting macros or plugins, please share them with the rest of the community!