An automated procedure for object classification in astronomical images is presented. We implement the use of Support Vector Machines as opposed to previous automated object classification efforts using neural networks to establish a high accuracy object identification method for astronomical objects in large scale astronomical images. Existing detection techniques combined with the use of segmentation for the extraction of descriptor features is used. Support Vector Machines are then utilised to distinguish and classify the extracted objects. Our results show potential for creating a high accuracy system for classifying astronomical objects and more specifically star/galaxy separation in large field astronomical images. The development and implantation of SVMs for distinguishing astronomical objects in images offers a new more accurate tool for the analysis of large field astronomical images.