Abstract: Disclosed subject matter relates generally to image processing that includes a method of automatically establishing a relationship between a plurality of User Interface (UI) elements. The present disclosure provides detailed insights related to each UI element in each input image of an application utilizing minimal human intervention. An image processing system identities the UI elements and positions of the UI elements in input images received from one or more sources. Further, a label corresponding to each UI element in each input image is obtained. Finally, a level of each UI element identified in each UI screen is determined to automatically establish the relationship between the UI elements across UI screens. The relationship is established automatically between the plurality of UI elements in real-time, that helps in easy navigation within the UI screen and among the UI screens of the application.