Fiberlyzer

Fiberlyzer is an open-source, image-based method for automated, quantitative, and standardized assessment of visual fibrousness in plant-based meat analogues. While mechanical texture analysis provides important insights into material properties, visual fibrous structures are often evaluated subjectively and rely heavily on expert judgment. Fiberlyzer addresses this gap by providing an objective computer vision pipeline for characterizing fibrous appearance directly from 2D images. The method segments fibrous regions from images and extracts fiber shape features that describe structural organization, including fiber length, width, area, and branching behavior. A key metric, the fiber score (length-to-width ratio), shows strong correlation with expert panel evaluations, particularly when comparing formulations. In addition to supporting formulation optimization, Fiberlyzer enables structural similarity analysis between different samples and reference materials, such as cooked chicken breast. With a simple imaging setup and user-friendly interface, Fiberlyzer is designed for integration into formulation development, quality control, and production workflows. By reducing dependence on expert visual inspection, the method offers a rapid, low-cost, and reproducible approach to evaluating visual fibrousness in meat analogues.