This project is the dataset used by the paper and some core code.In the paper we used three dataset.Datasets used in our paper include corn weed dataset, lettuce weed dataset and radish weed dataset.
Our corn weed dataset was taken from the natural environment of the corn seedlings field. The Canon PowerShot SX600 HS camera were used vertically towards the ground to acquire images which can reduce the sun light reflection influence. The dataset of corn weed includes 1200 pieces of bluegrasss, 1200 pieces of cirsium setosums, 1200 pieces of sedges, 1200 pieces of chenopodium albums and 1200 pieces of corns.
Our lettuce weed dataset was taken at the vegetable base in Shandong Province, China. The data collection lasts two different days and, in total 500 lettuce seedlings images and 300 weeds images were acquired from a 30cm height.
Our radish weed dataset is taken from public dataset.The dataset of radish weed included 200 radish seedlings and 200 associated weeds. The radish weed dataset from https://github.com/lameski/rgbweeddetection ,and the dataset is freely available for non-commercial applications,and this dataset contain 39 images with the same size of 3264 × 2448.Considering the original image numbers are small and image sizes are large, we cropped the original large images into smaller images. Finally, 200 images of radish seedlings and 200 images of its associated weeds were obtained and used in our experiments.
Here, we are very grateful to the dataset provided by Petre Lameski, which has greatly helped us in our experiments.
Our dataset will also be made public, and the dataset is freely available for non-commercial applications