• about reply
NET Reply UK Logo
Menu
  • About us
    About us
    • How we work
    • Our Leadership Team
  • Competences
    Competences
    • Networks
    • Future Networks
    • Security
  • Newsroom
  • Careers
  • Contact Us
    • about Reply
    NET Reply UK Logo
    Focus On

    Blog

    AI Image Process Automation & its Practical Uses

    FOCUS ON: artificial intelligence,

    WHAT IS IMAGE PROCESS AUTOMATION?

    In simplest terms, Image Process automation uses artificial intelligence (AI) to detect and label objects in images, the text is extracted and analysed using optical character recognition (OCR), bar codes can also be detected and decoded. Image processing is accomplished with an AI model and a collection (dataset) of images, used to train said AI model. Once trained, the AI model can identify text or objects within an image.

    The training process requires the annotation of images that are provided to the model as a training set. Future detections are done based on the experience gained from the training data. Therefore, the quality and size of the dataset will heavily affect the accuracy of the AI model: A dataset that has not been cleaned or annotated effectively will result in an inaccurate AI model.


    NET REPLY IMAGE PROCESSING TOOL

    Net Reply has developed an image processing tool that can be customised to meet a wide range of requirements. In its basic form, the tool takes an input of an image and returns the image with emphasised text that has been detected and the data that was extracted. The tool has been built with microservices architecture for future ease customisation and maintenance. A fully interactive dashboard comes with the Reply Image Processing tool that allows users to upload images and view data on mobile and desktop. An API endpoint is available If you decide that you don't need a dashboard and would like to integrate it with your software.

    AI Image Process Tool Steps

    Using our custom AI models, purpose-built depending on the customer requirement. We can train our models to detect and identify almost any object in any image.


    MAIN FEATURES

    • Detection and labelling of objects in images
    • Detection and analysis of text in images
    • Detection and decoding of bar codes
    • Front-end REST API for custom web services
    • Back-end REST API for AI model integration
    • Custom models to detect specialist objects
    • Process automation using AI model results
    • Support for mobile, desktop and server clients


    USE CASES

    The application has many use cases largely due to the use of custom AI models that are studied, trained, and refined in-house. Examples of other use cases include:

    • Standard Compliance in a Data centre
      • Validating cabling is neat and tidy
      • Validating racks are not over utilised
      • Validating equipment is not showing any alarms
      • Validating equipment has resilient power
      • Validating equipment is turned on
      • Validating equipment has been installed in the correct position
      • Validating equipment has been labelled properly (and exists in an inventory system)
      • Inventory/stock checks
        • Equipment inventory using text or barcode recognition
      • Tool identification
        • Engineers can check if they have the tools required for their next job
      • Sign language Recognition
        • Computer vision-based recognition of real-time or static hand gestures

      The application uses are not limited to the IT industry. For example, the inventory use case can be applied in many commercial settings: Take an inventory within a library, where taking a picture of a whole bookshelf and getting back data of each book can be done in an instant, compared to scanning each book individually, which would be a lengthy process.


      BENEFITS

      The tool’s benefits include large cost savings due to speeding up very manual and error-prone tasks. E.g., near real-time validation of field service installations preventing site re-visits, data centre compliance checks and audits, validating field service supplies…

      Therefore, any task which requires photos to be taken and processed can benefit greatly from this. There is a huge scope for further development in terms of viable use cases: This could be using the optical character recognition (OCR) capabilities of the tool for data extraction or training new models to recognise correct or incorrect states in a completely different scenario.

      What do you think would make a good use case for this technology?


    If you like the tool and would like to know more about it, or be given a demo, feel free to reach out to us or follow us on LinkedIn !

    RELATED CONTENTS

    07.04.2022 / Artificial Intelligence

    Blog

    AI Image Process Automation Tool - Demo

    In our previous blog, we introduced our AI Image Processing Tool, its capabilities and applications. This blog will now cover an example of a practical use case to provide you with food-for-thought. Read to discover a step-by-step application of our Tool!

     
     
     ​
     
    Reply ©��​ 2023​ - Company Information -
     PrivacyCookie Settings ​
    • About Reply​
    • Inves​tors​​​
    • Newsroom
    • Follow us on
    • ​
    ​
    • ​Privacy & Cookies Policy
    • Information (Client)
    • Information (Supplier)
    • Information (Candidate)
    • Modern Slavery Act Tran​sparency Statement (UK & IR)​
    ​​Reply Enterprise Social Network​