Abstract: High-purity quartz is a critical raw material for high-tech industries such as semiconductors, optical fiber communication, and laser technology. Its quality directly determines the performance and application efficiency of quartz-based products. Among the factors influencing its quality, submicron fluid inclusions within quartz particles not only affect the purity grade but also significantly contribute to bubble defects during material melting. However, traditional methods for quantifying fluid inclusions in quartz particles are limited by low efficiency and insufficient accuracy, making them inadequate for large-scale screening and precise evaluation of high-purity quartz raw materials. To address these challenges, this study proposes an automated identification algorithm based on color-space threshold segmentation. By capturing polarized dark-field and transmitted bright-field images of fluid inclusion slides under a polarizing microscope, the proposed method uses the YCbCr color channels of the polarized dark-field images to segment and extract quartz particles. Subsequently, the segmented results are overlaid with transmitted bright-field images to identify fluid inclusions within quartz particles and calculate their area ratio, enabling quantitative analysis of inclusion characteristics. Based on this algorithm, a software system with a B/S architecture is developed, integrating functionalities such as quartz particle feature extraction, fluid inclusion ratio calculation, and data visualization. To evaluate the performance of the software and algorithm, experiments are conducted to assess computation time and result stability under different sample capacities. The results show that computation time increases linearly with sample size, with larger sample capacities exhibiting greater fluctuations in computation time. Regarding stability, the pixel area of quartz particles is used as a reference metric. When the sample capacity exceeds 30 groups, the median and skewness of the particle size distribution curve tend to converge. In addition, validation experiments are conducted on two quartz samples with significantly different inclusion contents. The results demonstrate that the proposed method can accurately identify and quantify the area ratio of fluid inclusions within quartz particles, showing a high degree of consistency with manual observations. This study provides an efficient and accurate technical solution for the refined evaluation and screening of high-purity quartz, and it offers insights for the identification and quantitative analysis of inclusions in other mineral materials. Future work will focus on incorporating deep learning techniques to enhance the algorithm’s robustness and generalization ability.