Bark Classification of Trees Using K-Nearest Neighbor & Nearest Neighbor Algorithms

Pakistan is an agricultural country and less than 4 % of area secured with forests. Tree automatic classification based on computer science and it is the developing trend of classification. In this paper we examine how we can done bark classification of trees using k-nearest neighbor and nearest neighbor algorithms. There we discuss how these algorithms can be used to automatically classify trees from images of bark. We get the images of five kinds of different trees names suppose as A, B, C, D and E through using digital camera. We take ten different images of each kind of trees. The capability and information of inspectors are essential to perfectly achieve this process. The all the process will be done in computer vision image processing tool. In this tool we use the Histogram Features, Texture Features, and Pattern Classification. We achieved the final results of five kinds of different trees using nearest neighbor on distance two 82% average and on k-nearest neighbor when k=2 then the average result 82%, when k=3 the average result 82%, when k=4 then the average result 76% and when k=5 the average percentage 72% the result shows the maximum correct result and classifies the trees. These are the best percentage results using these algorithms for classification. In this way we can easily classify the different trees and also these methods provide opportunity to farmer and other people for identify and select the different better different trees for getting more benefit. Keywords: CVIP Tool, Histogram Features, Texture Features, Pattern Classification, Classification Algorithm