A Large-Scale Benchmark and Challenges for Object Detection in Aerial Images
We introduce two detection tasks. Task1 uses the initial oriented bounding boxes (OBB) as ground truth. Task2 uses the generated HBB as ground truth. Task1 has more practical value and we recommond you to test your algorithms in Task1.
The aim of this task is to locate the ground object instances with an OBB.
You need to submit a zip file containing results for all test images for evaluation. The results are stored in n (n is the numer of categories) files, "Task1_plane.txt, Task1_storage-tank.txt, ...", each file contains all the detections for a specific category. Each file is in the following format:
An example submission of task1 for DOTA-v1.0
imgname score x1 y1 x2 y2 x3 y3 x4 y4 imgname score x1 y1 x2 y2 x3 y3 x4 y4 ...
The evaluation protocol for oriented bounding box is a little different from the protocol in the original PASCAL VOC. We use the intersection over the union area of two polygons(ground truth and prediction) to calculate the IoU. The rest follows the PASCAL VOC.
Detecting object with horizontal bounding boxes is usual in many previous contests for object detection. The aim of this task is to accurately localize the instance in terms of horizontal bounding box with (xmin, ymin, xmax, ymax) format. In the task, the ground truths for training and testing are generated by calculating the horizontal bounding boxes over original annotated bounding boxes.
You need to submit a zip file containing results for all test images for evaluation. The results are stored in n (n is the numer of categories) files, "Task2_plane.txt, Task2_storage-tank.txt, ...", each file contains all the results for a specific category.
An example submission of task2 for DOTA-v1.0
imgname score xmin ymin xmax ymax imgname score xmin ymin xmax ymax ...
The evaluation protocol for horizontal bounding boxes follows the PASCAL VOC benchmark, which uses mean Average Precision(mAP) as metric.