KITTI数据集测试

来源：互联网发布：随机抽取软件pc 编辑：程序博客网时间：2024/06/06 19:50

KITTI image 序列转为rosbag 文件

利用python将kitti的数据转为rosbag, 仅测试于grayscale数据集,测试过程如下.
python代码请见:
http://git.oschina.net/taiping.z.git/image2rosbag_KITTIodometry

1. 在KITTI网页下载odometry dataset (grayscale, 22GB), 并解压.

2. 指定image路径, 生成的bag名,时间戳路径, 然后运行.

python img2bag_kitti_odo.py /your directory/KITTI/dataset/sequences/00/image_0 kitti_00_l.bag /your directory/KITTI/dataset/sequences/00/times.txt

3. 查看rosbag,并测试结果

查看rosbag

rosbag info image_0 kitti_00_l.bag--------path:        kitti_00_l.bagversion:     2.0duration:    7:50s (470s)start:       Jan 01 1970 08:00:00.00 (0.00)end:         Jan 01 1970 08:07:50.58 (470.58)size:        2.0 GBmessages:    4541compression: none [2271/2271 chunks]types:       sensor_msgs/Image [060021388200f6f0f447d0fcd9c64743]topics:      camera/image_raw   4541 msgs    : sensor_msgs/Image

在ORBSLAM2上测试

rosbag play --pause kitti_00_l.bagrosrun ORB_SLAM2 Mono Vocabulary/ORBvoc.txt Examples/Monocular/KITTI00-02.yaml

注意此处我们已经将rostopic设置为了camera/image_raw.

update Sep.23 2017

4. image to compressed image

压缩图片以便存储

 rosrun image_transport republish raw in:=(in_base_topic) compressed out:=(out_base_topic)

image_transport republic 压缩或者解压视频信息存储在rosbag.

阅读全文

1 0