The need to analyse big data streams and prescribe actions pro-actively is pervasive in nearly
every industry. As growth of unstructured data increases, using analytical systems to assimilate
and interpret images and videos as well as interpret structured data is essential. In this paper,
we proposed a novel approach to transform image dataset into higher-level constructs that
can be analysed more computationally efficiently, reliably and extremely fast. The proposed
approach provides a high visual quality result between the query image and data clouds with
hierarchical dynamically nested evolving structure. The results illustrate that the introduced
approach can be an effective yet computationally efficient way to analyse and manipulate storedimages
which has become the centre of attention of many professional fields and institutional
sectors over the last few years.