2016年使用精细语义分割网络的多物种水果花检测图像数据集

果实花朵数据集为检测树果实花朵所需的评估语义分割网络创建。图像使用彩色相机在天然果园条件下拍摄，涉及苹果、桃和梨三种果树花朵的图像数据。数据集中有四组图像，两组是苹果花。另外两个分别是桃花和梨花。整个数据集包含130张5184×3456像素的图像和60张2704×1520像素的图像。花朵的逐像素注释以单独的二进制图像集形式提供，初始通过徒手注释和区域生长细化执行。https://data.nal.usda.gov/dataset/data-multi-species-fruit-flower-detection-using-refined-semantic-segmentation-network

This fruit flower dataset was developed to evaluate semantic segmentation networks for tree fruit flower detection. The images were captured using color cameras under natural orchard environments, and the dataset encompasses image data of flowers from three fruit tree species: apple, peach, and pear. The dataset is split into four image groups: two groups contain apple flower images, while the remaining two groups correspond to peach flower and pear flower images respectively. The full dataset includes 130 images with a resolution of 5184 × 3456 pixels and 60 images with a resolution of 2704 × 1520 pixels. Pixel-wise annotations for the flowers are provided as separate binary image sets, which were initially created via manual annotation and refined using region growing techniques. https://data.nal.usda.gov/dataset/data-multi-species-fruit-flower-detection-using-refined-semantic-segmentation-network