Abstract
The production of high quality grapes for wine making is challenging. Significant progress has been made in the automated prediction of harvest yields from images but the analysis of images to predict the quality of the harvest has yet to be fully addressed. The quality of wine produced depends in part on the quality of the grapes harvested and therefore on the presence of disease in the vineyard. There is potential for automated early detection of disease in grape crops through the development of accurate techniques for image processing. This paper presents a review of current research and highlights some of the key challenges for geo-computation (image processing, computer vision and data mining techniques) to inform the management of vineyards and highlights the key challenges for in-field image capture and analysis. An exploration of potential applications for the knowledge generated by imaging techniques is then presented. This discussion is driven by the current interest in the effect of rapid and dramatic climate change on the production of wine and focuses on how this information might be utilized to inform the design and validation of accurate predictive models.
