Defect detection and quality control play an important role in the textile industry. In this paper, an automatic algorithm based on the optimal Gabor filter is proposed for real-time inspection of textile fabrics. The Cuckoo optimization algorithm is adopted to optimize the parameters of the Gabor filter. Also, an adaptive local binarization method is proposed, which enhances the performance of our algorithm. In order to locate the defects, the filtered image is divided into non-overlapping blocks. Then, the candidate defective blocks are binarized using adaptive thresholds, which are determined by blocks statistics. The performance of the proposed algorithm is evaluated through different types of fabrics in the TILDA database and an online Fabric Stain Dataset. The experimental results demonstrate the efficiency of the proposed method in detecting defects on the plain, regular and irregular patterned fabrics. Furthermore, the comparative results are provided to show the robustness of the proposed method.

缺陷检测和质量控制在纺织工业中起着重要作用。本文提出了一种基于最优Gabor滤波器的自动算法，用于织物的实时检测。采用杜鹃优化算法对Gabor滤波器的参数进行优化。同时，提出了一种自适应的局部二值化方法，提高了算法的性能。为了定位缺陷，将滤波后的图像划分为不重叠的块。然后，使用自适应阈值对候选缺陷块进行二值化，该阈值由块统计信息确定。通过在TILDA数据库中使用不同类型的面料和在线面料染色数据集来评估所提出算法的性能。实验结果证明了该方法在平纹，规则和不规则图案织物上检测缺陷的效率。此外，提供的比较结果表明了该方法的鲁棒性。