Cars Overhead with Context (COWC). In Lawrence Livermore National Laboratory (LLNL) Open Data Initiative

The Cars Overhead With Context (COWC) dataset is a large set of annotated cars from overhead. It is useful for training a device such as a deep neural network to learn to detect and/or count cars. The COWC dataset has the following attributes: 1. Data from overhead at 15 cm per pixel resolution at ground (all data is EO). 2. Data from six distinct locations: Toronto Canada, Selwyn New Zealand, Potsdam and Vaihingen Germany, Columbus and Utah United States. 3. 32,716 unique annotated cars. 58,247 unique negative examples. 4. Intentional selection of hard negative examples. 5. Established baseline for detection and counting tasks. 6. Extra testing scenes for use after validation. The data includes wide area imagery with annotations as well as precompiled image sets for training/validation of classification and counting. Examples of the precompiled image sets are provided. A newer subset (COWC-M) also differentiates between four different types of automobiles. a) Sedan b) Pickup c) Other d) Unknown

带上下文高空车辆（Cars Overhead With Context, COWC）数据集是一类大规模高空拍摄的带标注车辆数据集，可用于训练深度神经网络等模型，以实现车辆检测与计数任务。该数据集具备如下属性：1. 影像采集自高空场景，地面分辨率为15厘米/像素，所有数据均为EO影像；2. 数据覆盖6处独立采集区域：加拿大多伦多、新西兰塞尔温、德国波茨坦与瓦因根、美国哥伦布与犹他州；3. 包含32716个唯一标注车辆样本与58247个唯一负样本；4. 采用刻意筛选的难例负样本；5. 为车辆检测与计数任务确立了标准基准；6. 预留了额外的测试场景，可用于验证完成后的模型测试。该数据集涵盖带标注的大范围影像，以及用于分类与计数任务训练/验证的预编译影像集，预编译影像集的示例已随数据集一并提供。其新增的COWC-M子集可区分四类不同的汽车类型：a) 轿车；b) 皮卡；c) 其他车型；d) 未知类型。