The air bubbles defects in tires are one of the reasons that cause a tire explosion, which may endanger human life. Currently, the factories depend significantly on experienced operators for tire defect diagnosis process. This process requires a significant amount of experienced human resources, and this process takes time and requires a certain cost. With the development of technology, it has become possible to automate the classification process by using different techniques to reduce the possibility of human error. Digital shearography is a popular method to help detecting bubbles from images that are not normally visible with naked eye. In this study, we proposed a model to detect air bubble defects from shearography images of tires obtained from the Pirelli Automobile Tires Izmit factory. This study consists of two stages. In the first stage, three types of texture features are extracted from the shearography images. The feature set has 24 Gray Level Co-Occurrence Matrix (GLCM) features, and 16 Gray Level Run Length Matrix (GLRLM) features, 16 Gray Level Size Zone Matrix (GLSZM) features. Then, Multi-Layer Perceptron (MLP), Random Forest (RF), and Gradient Boosting classifiers are used for classification stage. The data set utilized in this study is 793 images without air bubbles and 207 images with air bubbles. The obtained results showed promising results for classification of air bubbles defects in tires.