Recognizing tree diseases at a distance via image processing and computer vision

• Main Author: Chan, Zhi Xian.
• Other Authors: School of Electrical and Electronic Engineering
• Format: Final Year Project
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/53262
• Institution: Nanyang Technological University

The project shows a method of recognizing different tree diseases based on distance via using image processing. The following technique is being used was by partitioning each image into sub regions and compute histograms of local features found inside individual sub regions. Any interesting points on the object can be extracted to provide a feature description of it. Hence, Bag-of-features model is used to represent as an unordered collection of words and it has been widely used for computer vision. However, Spatial Pyramid is a better extension of an orderless Bag-of-Words which shows significant improvement of performance. To perform recognition, the features extracted from the training image are feed into Support Vector Machine for data classification. Based on the data classification, we would be able to see which type of trees diseases on highest accuracy.