基于YOLOv3和brox光流的运动目标检测算法，对动态背景进行了运动补偿.zip

# KITTI-VOL-YOLO-format Some scripts for converting kitti data format to YOLO ## STEP1. modify_annotations_txt.py This script is used for **merging and modifying** labels of origin kitti labels. **Note:** backup the origin kitti labels before you use it ## STEP2. txt_to_xml.py This script is used for **generating VOL format labels** like *xxx.xml*. You should mkdir 'Annotations' folder in current dir before you use it ## STEP3. kitti_label.py This script is modified by 'voc_label.py' provided by ***pjreddie/darknet*** to generate labels for YOLO. Generated labels has the same name with the pictures and also the origin labels. It also generates *train.txt* which contains absolutely paths of the pictures, which is important for training on YOLO. **Note:** you should put the labels in the same folder of pictures before you train ## STEP4. create_train_test_txt.py ***(optional)*** Randomly divide kitti dataset into proportional parts like *train.txt* *test.txt* *val.txt* *trainval.txt* and so on. This *.txt* files contains absolutely paths of the pictures like **STEP3**, and can be directly used for training. Also you should create or modify relevant folders which you can see in the code before you use it