The BIT-Vehicle dataset contains 9,850 vehicle images. There are images with sizes of 1600*1200 and 1920*1080 captured from two cameras at different time and places in the dataset. The images contain changes in the illumination condition, the scale, the surface color of vehicles, and the viewpoint. The top or bottom parts of some vehicles are not included in the images due to the capturing delay and the size of the vehicle. There may be one or two vehicles in one image, so the location of each vehicle is pre-annotated. The dataset can also be used for evaluating the performance of vehicle detection. All vehicles in the dataset are divided into six categories: Bus, Microbus, Minivan, Sedan, SUV, and Truck. The number of vehicles per vehicle type are 558, 883, 476, 5,922, 1,392, and 822, respectively.
6 vehicle types, 9,850 images.
sedan [sɪ'dæn]：n. 轿车，轿子
Vehicle Type Classification Using a Semisupervised Convolutional Neural Network
Vehicle Type Classification Using Unsupervised Convolutional Neural Network
Z. Dong, Y. Wu, M. Pei and Y. Jia, “Vehicle Type Classification Using a Semisupervised Convolutional Neural Network,” in IEEE Transactions on Intelligent Transportation Systems, vol. 16, no. 4, pp. 2247-2256, Aug. 2015.