Custom capturing tool for extracting text from a scanned papers (e.g. Invoices)

We want to build a custom capturing tool for extracting text from a scanned papers (e.g. Invoices). So, My Question is How the OCR.Space APIs can help us to extract a specific zones (will be identified during the training phase) with specific dimensions (x,y,top,bottom) coordinates ?

The OCR API returns the bound box coordinates for every word. You can then find/exclude text based on the x/y area of your interest.