Fiji (ImageJ)

Fiji is a distribution of ImageJ that ships with the plugins, libraries and update sites a life scientist actually needs, under the only honest tagline in scientific software: “ImageJ, with batteries included”. ImageJ itself, originally written at the NIH, is the most widely used scientific image analysis package in biology, and Fiji is how most people should install it.

What it does

Fiji opens essentially any microscopy format through Bio-Formats, then provides the full stack of quantitative image analysis: segmentation, thresholding, particle counting, colocalisation, intensity measurement, deconvolution, registration, stitching, 3D rendering and tracking. TrackMate handles particle and cell tracking, Trainable Weka Segmentation brings machine-learning-based classification, and the update-site mechanism lets labs distribute their own plugin sets.

Everything is scriptable. The macro language is the low barrier to entry, and Jython, Groovy, JavaScript, Beanshell and Clojure are all available for anyone who wants real control. A recorded macro turns a manual click-through into a reproducible batch pipeline, which is the single most valuable habit an imaging lab can adopt.

Why it matters

Image quantification is where a great deal of biology quietly goes wrong. Manual thresholding chosen by eye, analysed on a handful of representative fields, is not a measurement. Fiji makes the alternative — a scripted, parameterised, batch-applied pipeline that treats every image identically — accessible without a programming background, and it costs nothing.

Strengths

• Free and open source, with no licence server and no per-seat cost.
• Bio-Formats reads over a hundred proprietary microscope formats.
• Vast plugin ecosystem covering most published image analysis methods.
• Fully scriptable and batchable, in several languages.
• Enormous community, extensive documentation, and a forum where method authors answer questions.
• Runs on Windows, macOS and Linux, including on modest hardware.

Limitations

The interface is dated and discoverability is poor — capability is buried in nested menus and plugin names that assume you already know what you are looking for. Memory management on very large datasets can be awkward, and Java heap errors are a rite of passage. For modern deep-learning segmentation and large multidimensional data, Python tools such as napari, Cellpose and QuPath have moved ahead in specific niches, though Fiji remains the general-purpose default.

Verdict

If you generate microscopy images, install Fiji rather than plain ImageJ and learn the macro recorder in your first week. It is free, it is the field standard, and the reproducibility gained from scripting your analysis is worth more than any commercial package you might otherwise buy.