Brands or trademarks have a great role in today’s marketing world. Companies, products, and various gaming leagues are recognized by their individual logos. Brand recognition in images is the main problem in a wide array of applications, like copyright infringement detection, vehicle branch detection for intelligent traffic-monitoring systems, placing contextual advertisements contextually, augmented reality etc. The brand detector addresses the need to identify brands in images with the help of Computer Vision. When an image containing an object whose brand is to be identified is uploaded, the brand detector identifies and displays the brand to the user along with the confidence level, thereby helping users ascertain the brand and further associate with it.
This project uses 3 Azure services-
1) Azure Cognitive Services
2) Computer Vision
3) Text to Speech
The app has been deployed to the URL- https://brandapp123.herokuapp.com/
The steps to use the app deployed on the working Project URL are-
1. Landing Page
2. Upload a picture of the image that you want to identify a brand in by clicking on the browse and then upload button
3. View the results which contain the brand name detected, the colour scheme of the brand and the audio version of the identified brand
To listen to the audio, ensure that the "Allow Audio" setting of your browser is on-
The project URL submitted works without any error and from all devices.
The screenshots have been taken from the deployed web app
Also, the Project as stated before uses 3 Azure services.
For checking which Azure services were used, kindly view the app.js file
Please view the Azure services being deployed in the app.js file
The node modules for this project contain the Azure dependencies as well. But, owing to the huge size of the node modules, I have not uploaded them on Github. Just to provide a sample of the Azure dependenices in the node modules of the code, kindly check the cognitiveservices-computervision folder
Another example of a brand being recognized-
