To develop a video / image Object & Optical Character Recognition Analysis Engine that can identify
from a still Image (Video captured image or Camera / Scanner or Fax captured Image) all the parts on
We need to utilize Machine Learning, Computer Vision and Deep Learning techniques along with
mathematical algorithms to optimize the detection of these on an image.
The identified objects and OCR text will be segmented using outline boxes and is called the Object/OCR
Layout Analysis Template (OCRLAT).
The OCRLAT creates an image with overlay outline boxes that can be edited and stretched to fit the
detected type and / or changed to a different type than our engine detected.
These outline boxes can also be split and / or removed from identifying and / or displaying in Output
Forms Image specific areas and / or minimize showing specific data when Object/OCR completed and
output data is pre-populated in purposely setup Data display Forms
Once the OCRLAT is used to scan the areas and the Data is provided as Output exporting the data to an
ODBC format, it will enable the user to:
• Download the data as a CSV file or other ODBC compliant format
• Download images as multimedia files
• View Images and Text output in a Form layout.
Whilst the download is quite self-explanatory, the Form layout needs some explanation.
Basically the output ODBC compliant text and image data can be recalled via a specifically laid out form
we call the Output Layout Form template (OUTLFT)
This Layout Form Template:
• is initially auto generated from the creation based on the layout detected in OCRLAT creation
• can be edited re-arranging text and object cells using Drag and Drop editor and even editing text
and images and this new layout prompts user to confirm to save to a New Form or overwrite
existing (or default) Form
• can manage any Text or Image changes on a Form can also update the ODBC Database file or
create a new ODBC file to update data, to enable a later Database compare changes
• Any OUTFLT can also be linked to >1 OCRLAT template e.g. Multiple Passports need different
OCR LAT Templates but output has many common fields such as name, passport number Date of
Birth, face picture, expiry date and any VISAs in the Passport
• Any edited text in Forms can update ODBC database and/or create new ODBC database
Once Form is linked to the OCR Layout Analysis Form, it is tested, output is checked and any tweaks are optimized.
Lastly the OCR Capture template (OCRLAT) and Form(s) are linked and uploaded as templates to the
Object/OCR mobile and windows apps
27 freelancers are bidding on average $31714 for this job
Hello I want to see the images with image Object & Optical Character. So I can know if I can do your project or not. I think that we can discuss this project in chat. Best
Hello we are team have full experience of OCR and image processing Please contact me I will provide best service and working You can check my working for your job. I am ready to discuss with you Thanks