It’s important to know what type of stains you’re facing before choosing the most appropriate pretreatment and washing method to use. So we are looking to automate the recognition of type of stain on the cloth.
Basic guidelines while data collection are explained below:
Source - 2 megapixels or above/1080p phone camera or DSC.
700 images per category (Day/Night environment settings)
(1-2) ft distance between cloth and camera, Stain should be recognizable to human eye.
Different Cloth types:
• Different Fabric Types: Equal Quantity
• Different cloth color: Equal Quantity
• Printed/Texture pattern: 40%
• Plain: 60%
Types of stain on cloths:
• Tea/ Coffee
• Mud/ Grass
• Oil/ Color Paint/ Grease
• Food Stains
Need annotations of stains in csv format.
We need images which are not hosted on web with specification as mentioned above.
• 1-2 weeks for sample Data (5 images per each category to validate image quality).
• 4-6 Weeks for full dataset.
- Stain Image collection, Vendor need to source the Image approx. 5000 and also prepare the metadata.
-each category approx 700 images
Pricing First case- kindly share the Pricing image sourcing and metadata. (Rates / Image).
3 freelances font une offre moyenne de 6667 $ pour ce travail
With Image processing projects, I have worked with. With Machine Learning and Deep Learning, I am having sound expertise. High-Performance AI Virtual Assistant - [login to view URL] && [login to view URL] 1. Plus