this is original colmap run forward-facing data,scale drift !!!!!!!!!!!!!!!!!!
实现需要补充的代码部分（用仓库里面的代码replace下面目录即可，注意代码里面的变量要自己找个地方定义）：
bundle_adjustment.cc 是在src/optim 目录下
incremental_mapper.cc 是在src/sfm 目录下
cost_functions.h 是在src/base 目录下
I add gps constraint in global bundle adjustment,obviously，the drift disappear！
I have simulated tunnels or no GPS, and my integrated GPS still performs well，like this!
1、ISPRS
https://www.isprs.org/education/benchmarks.aspx
2、sensefly
https://www.sensefly.com/education/datasets/
3、Kagaru
https://michaelwarren.info/docs/datasets/kagaru-airborne-stereo/
4、WHU
http://gpcv.whu.edu.cn/data/WHU_MVS_Stereo_dataset.html
5、KITTI
http://www.cvlibs.net/datasets/kitti/
6、ETH
https://www.eth3d.net/
7、realitycapture
https://www.capturingreality.com/SampleDatasets
8、视觉定位数据集
https://www.visuallocalization.net/datasets/
9、ODM
https://www.opendronemap.org/odm/datasets/
10、mve 官网的数据
https://www.gcc.tu-darmstadt.de/home/proj/ambient_point_clouds/apc.en.jsp
11、奥地利维也纳理工大学的Harvest4D项目(有GT）
https://harvest4d.org/?page_id=1367
12、自动化所偏古建筑
http://vision.ia.ac.cn/zh/data/index.html
13、blendedmvs
https://github.com/YoYo000/BlendedMVS
14、Drone Mapping
https://dronemapper.com/sample_data/
16、https://www.cs.cornell.edu/projects/bigsfm/#code bigsfm
17、http://brian.harrison.org/Photogrammetry/JemezHistoricalSite/BodyJemez.html
Last, thank you very much for the excellent work of the author(@ahojnnes) of colmap.Later,I will add feature scale constraint in bundle adjustment!!!!!