A framework to estimate the water content in leaves by image processing and data mining

Identification of dryness of plants and trees is an important research area as it may help in predicting natural bushfires and planning crop watering in areas of water shortage. Due to the global climate changes the frequency of bushfires and scarcity of water for crop production has increased in many parts of the world. In order to estimate the water content in leaves two methods are employed which can be broadly categorised as invasive and non-invasive methods. Non-invasive methods are preferable as they do not need the removal of leaves from the plants. This research proposes an image processing and data mining technique to estimate the water content in leaves. The leaves will be identified from their background in the high resolution colour images and the features will be extracted. The extracted features will then be used to classify the leaves into different classes of leaf water content. The paper discusses the existing important leaf water content determining methods, their drawbacks in the use of current applications, and the proposed research objectives and the methodology.