The OpenSprayerSeg dataset is a meticulously curated dataset specifically designed for the field of image segmentation in agricultural applications. Originating from the Open Sprayer Dataset, this collection features 4172 RGB images, each with a resolution of 256x256x3 pixels. These images predominantly showcase weeds and are complemented by corresponding binary masks of the same resolution (256x256x1). The binary masks are designed with a dual classification system: '0' signifies crop pixels, while '1' denotes weed pixels.
The Open Sprayer Dataset, the parent source, contains 1173 original images featuring weeds. To expand this collection and enrich the dataset, various augmentation techniques were applied, including horizontal flipping, vertical flipping, and rotation. This process was instrumental in extending the dataset to its current volume of 4172 samples.
The binary masks within this dataset were crafted using Gimp, a renowned free and open-source image manipulation program.
You can access the OpenSprayerSeg data on Kaggle.
