Front-end Logistics application to manage inflow of stock, using google machine learning based vision and data modeling. This application contains
- A barcode scanner which automatically uses the google api crop_hints and ml_kit to analyze and parse various forms of barcores
- Failing that, defualts to a rudimentary scan of the book cover and matches to the appropriate book from a database using 4 factors. i) Euclidian Distance between text features and the standard logo; ii) Latin Text detected; iii)Labels identified by mk_kit; iv) The average color of the cover, seperated into 16 blocks; iv) Characters identified by custom machine learning model
- Failing that, goes to the manual entry of a 4 digit code corresponding to the stock item with the appropriate verification and validation that is required.
Data Model was trained with 4 labels in tensorflow.Lite