The objective of this work is to explore the capabilities of the Julia programming language for image segmentation tasks in the field of microscopic plant analysis. We focus on the segmentation of key plant components in microscopic images of herbal mixtures. The study outlines a data preparation pipeline implemented entirely in Julia, including noise removal, contrast adjustment, and binarization, leveraging Julia’s native packages such as Images.jl and ImageFiltering.jl. We then apply a U-Net convolutional network, implemented using Flux.jl, to perform semantic segmentation. The results demonstrate that Julia can serve as an efficient and fully integrated toolchain for image analysis and machine learning workflows.