Effect of pan-sharpening multi-temporal Landsat 8 imagery for crop type differentiation using different classification techniques

• Machine learning can be employed to accurately differentiate between crops (∼95%).
• Pan-sharpening Landsat 8 imagery dramatically improves crop classification accuracy (∼15%).
• Pan-sharpening Landsat 8 imagery effects classification accuracy more than image analysis method does.

This study evaluates the potential of pan-sharpening multi-temporal Landsat 8 imagery for the differentiation of crops in a Mediterranean climate. Five Landsat 8 images covering the phenological stages of seven major crops types in the study area (Cape Winelands, South Africa) were acquired. A statistical pan-sharpening algorithm was used to increase the spatial resolution of the 30 m multispectral bands to 15 m. The pan-sharpened images and original multispectral bands were used to generate two sets of input features at 30 and 15 m resolutions respectively. The two sets of spatial variables were separately used as input to decision trees (DTs), k-nearest neighbour (k-NN), support vector machine (SVM), and random forests (RF) machine learning classifiers. The analyses were carried out in both the object-based image analysis (OBIA) and pixel-based image analysis (PBIA) paradigms. For the OBIA experiments, three image segmentation scenarios were tested (good, over and under segmentation). The PBIA experiments were carried out at 30 m and 15 m resolutions. The results show that pan-sharpening led to dramatic (∼15%) improvements in classification accuracies in both the PBIA and OBIA approaches. Compared to the other classifiers, SVM consistently produced superior results. When applied to the pan-sharpened imagery SVM produced an overall accuracy of nearly 96% using OBIA, while PBIA’s overall accuracy was 1.63% lower. We conclude that pan-sharpening Landsat 8 imagery is highly beneficial for classifying agricultural fields whether an object- or pixel-based approach is used.