In the 1990s, the CBIR method was proposed to overcome
the difficulties encountered with text-based image retrieval.
There are three methods of CBIR – tree-based index,
hash-based index, and visual-word-based on inverted
index.
Tree-based index was proposed by Silpa-Anan and
Hartley [23]. They created multiple KD-trees along with
a same dataset to improve the performance. Although
the tree-based indexing structure can successfully divide
data space into a hierarchical tree structure, when the