ai-grid-hackathon

Green Space Detection

Keywords: Computer Vision, Image Segmentation

Important There is a nice presentation from DVZ (Dataprocessing Center; Deutsch: Datenverarbeitungszentrum) about the case with all information here

Green spaces in cities are vital for environmental health and quality of life. However, identifying and analyzing these areas, especially from aerial images, poses a challenge due to the complexity of urban landscapes.

The goal is to develop an automated solution that accurately detects and segments green spaces containing vegetation in aerial photographs of cities. This will help urban planners and environmentalists monitor and manage green spaces more effectively.

Most importantly, DMZ is interested in the following questions:

• Can the percentage of green coverage in the urban area be detected?
• Are there differences in the quality of the green spaces?
• Can potential areas for improving green coverage be identified?

The Dataset

Information about the dataset can be found here

Further Reading & Information

• https://opencv.org/
• https://www.tensorflow.org/tutorials/images/segmentation
• https://pyimagesearch.com/2018/09/03/semantic-segmentation-with-opencv-and-deep-learning/