In this thesis, we implement a 2 dimensional object recognition algorithm using multiple binary decision tree classifier which can be generated by a simple method. Only one feature is utilized to construct a binary decision tree hence total number of trees is the same as the number of features. The implemented algorithm consists of feature extraction part, decision tree constructing part and decision tree searching part. The decision tree searching algorithm consists of three methods. The first tree searching algorithm is named. "All Tree Searching (ATS)." ATS algorithm searches all decision trees. The second tree searching algorithm, called "Static Sequential Tree Searching (SSS)", searches the decision trees according to pre-defined searching order until the stopping coast is less than the continuing cost. The last tree searching algorithm is named "Dynamic Sequential Tree Searching (DSS)". DSS algorithm searches the decision trees with dynamic searching order of BDT. ATS algorithm shows the best performance in terms of recognition error rate but the worst in the case of recognition time. DSS algorithm has shortest recognition time. The above three algorithms show low error rate (0.3 % - 2.7 %) in the general shaped 2 dimensional object recognition experiments.