Ensuring the quality and safety of apples is vital in the agriculture and food industries, particularly in detecting internal fruit rot, a common and often invisible defect. Traditional inspection methods can be inefficient and imprecise, prompting the exploration of advanced technologies such as machine learning for more accurate detection. This study presents a machine learning-based approach for assessing the quality of apples, specifically identifying the presence of internal fruit rot. A total of 116 apples were analyzed; for training, 43 apples with internal fruit rot and 50 fresh apples were used, and for testing, 11 apples with internal fruit rot and 12 fresh apples were set aside. X-ray images of the fruits were obtained, followed by the segmentation of the region of interest to isolate the affected areas. Gabor feature extraction was applied to enhance texture representation, and t-distributed Stochastic Neighbor Embedding (t-SNE) was used for dimensionality reduction to facilitate effective classification. The AdaBoost algorithm was employed for classification, with a 10-fold cross-validation approach being used during training. The training results demonstrated an accuracy of 94.6%, a precision of 96.2%, a recall of 91.3%, and an F1 score of 93.5%. Isotonic Regression was applied to calibrate the model, ensuring the predicted probabilities accurately reflected the likelihood of internal rot. The model was then tested on the separate test set, where it achieved an accuracy of 91.4%, a precision of 92.8%, a recall of 88.3%, and an F1 score of 90.2%. These results highlight the potential of machine learning techniques, particularly when combined with advanced imaging and feature extraction methods, to enhance the accuracy and efficiency of quality inspection processes in the fruit industry. This study suggests that such approaches could be widely adopted for non-destructive quality assessment, leading to better product safety and consumer satisfaction.