<p>The measurement of heat-set frieze yarn crimp is currently based largely on manual methods that depend heavily on operator skill and involve a high level of labor intensity. As a result, these methods suffer from a lack of precision due to the complex three-dimensional nature of yarn crimp. Moreover, they do not support integration into online manufacturing processes. These limitations lead to inefficiencies in quality measurement and increased production costs. The vision of this research is to develop a new fractal-based index using image processing and fractal dimension analysis to quantify and classify yarn crimp, with the goal of integrating it into online quality control systems. Two- and three-dimensional images of yarn test specimens were captured using a digital camera, then processed and evaluated to determine their fractal dimensions using advanced algorithms in the MATLAB software toolbox. The results of this stage led to the proposal of an innovative index for determining the level of crimp, accurately describing the geometric complexity of the yarn. To verify the reliability and practical application of the proposed index, comparisons were made with manual evaluations performed by expert textile engineers. The findings demonstrated a high correlation with manual results (correlation coefficient of 0.96), and the index’s accuracy was confirmed in over 95% of cases. Beyond accurately identifying the level of yarn crimp, this method enables implementation in online production lines and offers an innovative framework for improving quality control, reducing costs, and increasing productivity in the textile industry.</p> Graphical abstract <p></p>