Efficiency Evaluation of C-Tree and KD-Tree in Content-Based Image Retrieval

The image retrieval problem is performed on the C-Tree and KD-Tree structures which have brought many positive results regarding retrieval time and precision. The C-Tree structure is built according to the data clustering method, while the KD-Tree is built according to the multi-layer data classification method. Therefore, the common goal of these two structures applies to the image retrieval problem with quite high efficiency. In this paper, the results obtained from the two C-Tree and KD-Tree structures are evaluated, analyzed, and compared to the image retrieval problem. The experiments are conducted on the same COREL and WANG image data sets to serve as a basis for evaluating the performance of these two structures together; At the same time, the results are also compared with other works to demonstrate the effectiveness of the experimental method. Finally, some disadvantages of each structure are also analyzed for further improvements, and these two structures are combined to propose an image retrieval model to improve accuracy based on the advantages of the C-Tree and KD-Tree.