The emergence of plant rust has proven to be a serious threat to agricultural yields. The identification of plant diseases is traditionally done through visual inspection that often lead to delays and significant amount of loss by the time diagnosis is made. On the other hand, advanced deep learning models perform well with accuracy but their large size of datasets and computational complexities make them unsuitable for deployment which limits their real time applications. This study focuses on a method to identify rust using both ground and aerial photographs of plants. It follows a sequence of steps, including pre-processing, extracting Hue, Saturation, and Value (HSV) color features, utilizing Gray-Level Co-occurrence Matrix (GLCM) for extracting texture features, combining these extracted features, and classifying the healthy and rust-affected crops. Machine learning based Random Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbor (KNN) models are used to achieve higher performance, resulting in greater accuracy in identifying rust pathogens. The performance of our model exceeds existing approaches and is also the highest among feature-based methods, making it the most potential method for rust detection.