Detection of microscopic fission track (FT) star-shaped clusters, developed in SSNTD by etching, created by fission fragments emitted from fissile particles irradiated by neutrons, is a key technique in nuclear forensics and safeguards investigation. It involves scanning and imaging a large area, typically 1-2 sq.cm, of a translucent SSNTD (e.g. polycarbonate sheet, mica, etc.) to identify the FT clusters, sparse as they may be, that must be distinguished from dirt and other artefacts present in the image. This task, if done manually, is time consuming, operator dependent, and prone to human errors. To solve the problem, an automated workflow have been developed for (a) scanning large area detectors, in order to acquire large images with adequate high resolution, and (b) an image processing scheme, implemented in ImageJ, to automatically detect the FT clusters. The scheme combines intensity-based segmentation approaches, a morphological algorithm capable of detecting and counting endpoints in putative FT clusters and thus enables rejection of non-FT artefacts. In this paper, the methodology is described and first very promising results shown.