Windows Segmentation Dataset
The "Windows Segmentation Dataset" is specifically compiled for the manufacturing sector, focusing on the production and quality control of window units. It consists of internet-collected images with a resolution spectrum from 150 x 150 to 1160 x 2120 pixels. The dataset is designed for semantic segmentation and bounding box tasks, encompassing a variety of window designs and styles.
Sample


Specification
Dataset ID
MD-Image-034
Dataset Name
Windows Segmentation Dataset
Data Type

Volume
About 40.9k
Data Collection
Internet collected images. Resolution ranges from 150 x 150 to 1160 x 2120
Annotation
Semantic Segmentation,Bounding Box
Annotation Notes
The dataset includes multiple windows
Application Scenarios
Manufacturing
Data Collections
This dataset contains approximately 40.9k images, each featuring different window configurations. The annotations not only identify the windows but also provide detailed segmentation and bounding boxes, facilitating precise analysis and inspection in window manufacturing processes. This extensive collection supports advancements in automated quality assessment, design evaluation, and manufacturing efficiency in the window production industry.
Quality Assurance

Related products
Any further information, please contact us.