Frequently Asked Questions — COLMAP 3.7 documentation
COLMAP已知相机内外参数重建稀疏/稠密模型 - thronsbird - 博客园
Colmap根据相机内外参数重建稀疏模型_m0_47677188的博客-CSDN博客_colmap 命令行
准备 images.txt
现利用 colmap的script 提取database中的 id-name 对应关系
colmap/scripts/python/database_slz.py
python /colmap/scripts/python/database_slz.py --database_path 04/model/database.db然后 通过 home_move/dataset_home/kitti/colmap0109/kitti_pose2colmap_images.py 创建images.txt
python kitti_pose2colmap_images.py ../04/04.txt id-name.txt 代码为
from scipy.spatial.transform import Rotation as R
import numpy as np
import sys
path_cam= sys.argv[1]
path_idname= sys.argv[2]
f= open(path_cam, 'r')
lines = f.readlines()
f_id_name= open(path_idname, 'r')
f_id_name_lines= f_id_name.readlines()
f_w= open('04/model/images.txt', 'w')
id_name= {}
for l in f_id_name_lines:
l= l.strip().split(' ')
id_name[int(l[1].split('.')[-2])]= int(l[0])
for i, line in enumerate(lines):
line= list(map(float, line.strip().split()))
rotation_matrix= np.array([[line[0], line[1], line[2]], [line[4], line[5], line[6]], [line[8], line[9], line[10]]])
r = R.from_matrix(rotation_matrix)
rquat= r.as_quat() # The returned value is in scalar-last (x, y, z, w) format.
# print('before ', rquat)
rquat[0], rquat[1], rquat[2], rquat[3] = rquat[3], rquat[0], rquat[1], rquat[2]
# print('after ', rquat)
out= np.concatenate( (rquat, np.array([line[3], line[7], line[11]])), axis= 0)
# print('out \n', out)
# input()
id= id_name[i]
f_w.write('%d '% (id))
f_w.write(' '.join([str(a) for a in out.tolist()] ) )
f_w.write(' 1 %06d.png'% i)
f_w.write('\n\n')
f_w.close()
执行 检测和匹配的步骤 安装前面的链接
执行 三角化
nohup colmap/build/src/exe/colmap point_triangulator --database_path 04/model/database.db --image_path ../04/image_2/ --input_path 04/model/ --output_path 04/triangulated/sparse/ > 04/0111-triangulate.log 2>&1 & 结果： 三角化的点数小于500 ， 可以看到问题比较大
可视化问题 —— 已经重装 colmap 3.5解决：
BA只持续很短，可视化的模型 还是像一个 枣核 ，并且点很少
此外 colmap 始终不绘制 相机的轨迹
同样的步骤进行 kitti 07的数据， 并且提取特征时 直接 设置 single-model，参数等（中间不再像参考文章那样修改内参） ，脚本如下
# conda env_36
# nohup sh run.sh &
rm nohup.out
rm 07/* ;
rm -rf 07/model/ ;
rm -rf 07/triangulated/ ;
mkdir -p 07/triangulated ;
mkdir -p 07/model ;
/home/shilinzhe/colmap/build/src/exe/colmap feature_extractor --ImageReader.camera_model SIMPLE_PINHOLE --ImageReader.single_camera 1 --ImageReader.camera_params 707.0912,601.8873,183.1104 --SiftExtraction.use_gpu 0 --database_path 07/model/database.db --image_path ../07/image_0/ > 07/feat-0112.log 2>&1 ;
python /home/shilinzhe/colmap/scripts/python/database_slz.py --database_path 07/model/database.db --output 07/id-name.txt ;
python kitti_pose2colmap_images.py ../07/07.txt 07/id-name.txt 07/model/images.txt;
touch 07/model/points3D.txt ;
cp /data1/shilinzhe/home_move/dataset_home/kitti/colmap0109/04/model/cameras.txt 07/model/ ;
/home/shilinzhe/colmap/build/src/exe/colmap exhaustive_matcher --database_path 07/model/database.db --SiftMatching.use_gpu 0 > 07/0112-match.log 2>&1 ;
/home/shilinzhe/colmap/build/src/exe/colmap point_triangulator --database_path 07/model/database.db --image_path ../07/image_0/ --input_path 07/model/ --output_path 07/triangulated/ > 07/0112-triangulate.log 2>&1 ;
出来的结果：
有轨迹，有点云 ，但是 比较抖动，回环 不好
下一步 考虑加上 回环 和 修改特征匹配方式： 顺序匹配， 以及设置不优化真值 pose
只有把 带真值位姿的 全量的 kitti 重建 ok了，才好说 不带真值位姿是否可以，自己的数据需要什么样的pose， 数据采样频率， 回环等等
0121更新;
上述bash 的基础上，把最后的部分采用 mapper 进行重建 ， 不用input 也就是不用 预先设置的上那个txt
/home/shilinzhe/colmap/build/src/exe/colmap mapper --database_path 07/model/database.db --Mapper.ba_refine_focal_length 0 --Mapper.ba_refine_extra_params 0 --image_path ../07/image_0/ --output_path 07/mapper_seq/ > 07/0120-seq-mapper.log 2>&1 ;
得到的结果：
过程输出截取
==============================================================================
Global bundle adjustment
==============================================================================
iter cost cost_change |gradient| |step| tr_ratio tr_radius ls_iter iter_time total_time
0 7.807499e+05 0.00e+00 3.87e+16 0.00e+00 0.00e+00 1.00e+04 0 1.89e+00 1.43e+01
1 7.742615e+05 6.49e+03 6.25e+15 3.02e+00 1.01e+00 3.00e+04 66 1.95e+01 3.38e+01
2 7.732063e+05 1.06e+03 5.52e+15 3.39e+00 1.02e+00 9.00e+04 100 2.66e+01 6.05e+01
3 7.723630e+05 8.43e+02 6.09e+15 4.63e+00 1.06e+00 2.70e+05 100 2.67e+01 8.71e+01
4 7.718831e+05 4.80e+02 1.39e+16 3.54e+00 1.04e+00 8.10e+05 100 2.85e+01 1.16e+02
5 7.717216e+05 1.61e+02 6.75e+15 1.81e+00 9.56e-01 2.43e+06 100 2.66e+01 1.42e+02
6 7.716013e+05 1.20e+02 6.73e+15 1.30e+00 9.38e-01 7.29e+06 100 2.90e+01 1.71e+02
7 7.714913e+05 1.10e+02 6.71e+15 1.19e+00 9.35e-01 2.13e+07 100 2.62e+01 1.97e+02
8 7.713880e+05 1.03e+02 1.13e+16 1.06e+00 9.42e-01 6.38e+07 100 2.68e+01 2.24e+02
9 7.712936e+05 9.44e+01 2.50e+16 1.06e+00 8.95e-01 1.26e+08 100 2.75e+01 2.52e+02
10 7.711956e+05 9.80e+01 7.21e+15 9.50e-01 9.29e-01 3.40e+08 100 2.80e+01 2.80e+02
11 7.711040e+05 9.16e+01 1.88e+16 9.81e-01 8.97e-01 6.79e+08 100 2.80e+01 3.08e+02
12 7.710125e+05 9.15e+01 2.32e+16 8.96e-01 8.86e-01 1.26e+09 100 2.68e+01 3.34e+02
13 7.709165e+05 9.60e+01 2.35e+15 9.31e-01 9.39e-01 3.77e+09 100 2.74e+01 3.62e+02
14 7.708293e+05 8.72e+01 2.18e+16 8.69e-01 9.02e-01 7.82e+09 100 2.96e+01 3.91e+02
15 7.707410e+05 8.84e+01 1.89e+16 8.99e-01 9.03e-01 1.65e+10 100 2.81e+01 4.20e+02
16 7.706498e+05 9.12e+01 7.60e+15 8.63e-01 9.14e-01 3.79e+10 100 2.52e+01 4.45e+02
17 7.705634e+05 8.64e+01 7.69e+15 8.78e-01 9.08e-01 8.28e+10 100 2.98e+01 4.74e+02
18 7.704759e+05 8.75e+01 2.57e+16 8.67e-01 8.78e-01 1.46e+11 100 2.92e+01 5.04e+02
19 7.703836e+05 9.23e+01 3.12e+15 8.66e-01 9.43e-01 4.37e+11 100 2.65e+01 5.30e+02
20 7.702979e+05 8.56e+01 4.44e+15 8.74e-01 9.14e-01 1.01e+12 100 2.58e+01 5.56e+02
21 7.702155e+05 8.24e+01 6.17e+15 8.59e-01 8.78e-01 1.79e+12 100 2.72e+01 5.83e+02
22 7.701279e+05 8.76e+01 7.75e+15 8.86e-01 9.07e-01 3.90e+12 100 2.86e+01 6.12e+02
23 7.700435e+05 8.44e+01 3.76e+15 8.57e-01 8.99e-01 7.94e+12 100 2.69e+01 6.39e+02
24 7.699559e+05 8.76e+01 8.89e+15 9.06e-01 9.05e-01 1.70e+13 100 2.70e+01 6.66e+02
25 7.698795e+05 7.64e+01 3.86e+16 8.64e-01 8.14e-01 2.25e+13 100 2.86e+01 6.94e+02
26 7.697868e+05 9.27e+01 9.37e+15 9.30e-01 8.88e-01 4.22e+13 100 2.85e+01 7.23e+02
27 7.697021e+05 8.47e+01 9.70e+15 8.75e-01 8.77e-01 7.40e+13 100 2.96e+01 7.52e+02
28 7.696174e+05 8.47e+01 2.24e+16 9.59e-01 8.62e-01 1.20e+14 100 2.82e+01 7.81e+02
29 7.695282e+05 8.91e+01 4.04e+15 8.91e-01 9.10e-01 2.66e+14 100 2.78e+01 8.09e+02
30 7.694412e+05 8.71e+01 6.05e+15 9.94e-01 9.18e-01 6.42e+14 100 2.67e+01 8.35e+02
31 7.693662e+05 7.49e+01 5.39e+15 8.14e-01 9.40e-01 1.92e+15 100 2.89e+01 8.64e+02
32 7.691876e+05 1.79e+02 1.25e+16 1.79e+00 9.37e-01 5.76e+15 100 2.81e+01 8.92e+02
33 7.690154e+05 1.72e+02 2.47e+16 1.28e+00 9.22e-01 1.00e+16 100 2.99e+01 9.22e+02
34 7.689903e+05 2.51e+01 7.63e+15 4.26e-01 7.25e-01 1.00e+16 100 2.80e+01 9.50e+02
35 7.689709e+05 1.94e+01 2.02e+16 4.09e-01 6.29e-01 1.00e+16 100 2.87e+01 9.79e+02
36 7.689315e+05 3.95e+01 3.02e+15 4.80e-01 8.35e-01 1.00e+16 100 2.79e+01 1.01e+03
37 7.688969e+05 3.46e+01 5.78e+15 4.08e-01 8.11e-01 1.00e+16 88 2.47e+01 1.03e+03
38 7.688970e+05 -6.77e-02 0.00e+00 7.69e-03 -8.36e-03 5.00e+15 3 2.17e+00 1.03e+03
39 7.688962e+05 7.17e-01 1.35e+16 7.80e-03 8.83e-02 3.21e+15 3 4.31e+00 1.04e+03
40 7.689004e+05 -4.22e+00 0.00e+00 2.15e-03 -5.23e-01 1.60e+15 3 2.35e+00 1.04e+03
41 7.688975e+05 -1.24e+00 0.00e+00 2.16e-03 -1.54e-01 4.01e+14 3 2.50e+00 1.04e+03
42 7.688991e+05 -2.94e+00 0.00e+00 2.16e-03 -3.64e-01 5.01e+13 3 2.59e+00 1.05e+03
43 7.688992e+05 -3.04e+00 0.00e+00 2.16e-03 -3.77e-01 3.13e+12 3 2.25e+00 1.05e+03
44 7.688973e+05 -1.10e+00 0.00e+00 2.16e-03 -1.36e-01 9.79e+10 3 2.26e+00 1.05e+03
45 7.688971e+05 -9.02e-01 0.00e+00 2.16e-03 -1.12e-01 1.53e+09 3 2.79e+00 1.05e+03
46 7.688971e+05 -9.31e-01 0.00e+00 2.16e-03 -1.15e-01 1.20e+07 3 3.03e+00 1.06e+03
47 7.688971e+05 -8.82e-01 0.00e+00 2.16e-03 -1.09e-01 4.67e+04 3 2.88e+00 1.06e+03
48 7.688976e+05 -1.39e+00 0.00e+00 2.10e-03 -1.73e-01 9.12e+01 3 2.15e+00 1.06e+03
49 7.688949e+05 1.32e+00 8.50e+15 3.71e-04 1.66e-01 7.03e+01 3 4.47e+00 1.07e+03
50 7.688956e+05 -7.20e-01 0.00e+00 3.29e-04 -1.01e-01 3.51e+01 2 2.72e+00 1.07e+03
Bundle adjustment report
------------------------
Residuals : 4888438
Parameters : 849539
Iterations : 51
Time : 1068.61 [s]
Initial cost : 0.399642 [px]
Final cost : 0.396596 [px]
Termination : No convergence
=> Completed observations: 1960
=> Merged observations: 1983
=> Filtered observations: 640
=> Changed observations: 0.001875
==============================================================================
Global bundle adjustment
==============================================================================
iter cost cost_change |gradient| |step| tr_ratio tr_radius ls_iter iter_time total_time
0 7.725322e+05 0.00e+00 8.66e+15 0.00e+00 0.00e+00 1.00e+04 0 2.01e+00 1.65e+01
1 7.643204e+05 8.21e+03 4.77e+15 3.92e+00 1.01e+00 3.00e+04 65 2.02e+01 3.67e+01
2 7.632501e+05 1.07e+03 9.31e+15 2.85e+00 1.00e+00 9.00e+04 100 2.92e+01 6.59e+01
3 7.627408e+05 5.09e+02 1.75e+16 2.37e+00 9.85e-01 2.70e+05 100 2.84e+01 9.43e+01
4 7.625455e+05 1.95e+02 5.73e+15 1.53e+00 9.66e-01 8.10e+05 100 2.85e+01 1.23e+02
5 7.624592e+05 8.63e+01 6.42e+15 9.56e-01 9.19e-01 1.97e+06 100 2.89e+01 1.52e+02
6 7.623833e+05 7.59e+01 2.42e+16 7.89e-01 8.90e-01 3.74e+06 100 2.86e+01 1.80e+02
7 7.623082e+05 7.51e+01 1.05e+16 9.49e-01 8.94e-01 7.35e+06 100 2.79e+01 2.08e+02
8 7.622373e+05 7.09e+01 8.35e+15 8.56e-01 8.78e-01 1.29e+07 100 2.83e+01 2.37e+02
9 7.621685e+05 6.88e+01 2.11e+16 9.78e-01 8.56e-01 2.02e+07 100 2.90e+01 2.66e+02
10 7.620950e+05 7.35e+01 8.14e+15 8.84e-01 9.11e-01 4.56e+07 100 2.76e+01 2.93e+02
11 7.620300e+05 6.50e+01 1.37e+16 9.90e-01 8.73e-01 7.81e+07 100 2.87e+01 3.22e+02
12 7.619597e+05 7.04e+01 4.58e+15 9.00e-01 9.17e-01 1.86e+08 100 2.87e+01 3.50e+02
13 7.618950e+05 6.46e+01 1.01e+16 9.93e-01 9.00e-01 3.82e+08 100 2.70e+01 3.77e+02
14 7.618326e+05 6.25e+01 9.89e+15 9.07e-01 8.57e-01 6.01e+08 100 3.02e+01 4.08e+02
15 7.617710e+05 6.16e+01 2.29e+16 9.90e-01 8.27e-01 8.35e+08 100 2.59e+01 4.33e+02
16 7.617012e+05 6.97e+01 9.98e+15 9.13e-01 9.05e-01 1.78e+09 100 2.81e+01 4.62e+02
17 7.616377e+05 6.35e+01 5.26e+15 9.86e-01 9.02e-01 3.68e+09 100 2.61e+01 4.88e+02
18 7.615769e+05 6.08e+01 9.99e+15 9.16e-01 8.68e-01 6.14e+09 100 2.72e+01 5.15e+02
19 7.615152e+05 6.17e+01 1.86e+16 9.81e-01 8.55e-01 9.57e+09 100 2.59e+01 5.41e+02
20 7.614473e+05 6.79e+01 3.80e+15 9.17e-01 9.29e-01 2.60e+10 100 2.53e+01 5.66e+02
21 7.613864e+05 6.09e+01 9.71e+15 9.72e-01 9.24e-01 6.65e+10 100 2.87e+01 5.95e+02
22 7.613268e+05 5.95e+01 3.28e+15 9.18e-01 8.88e-01 1.24e+11 100 2.93e+01 6.24e+02
23 7.612656e+05 6.12e+01 4.25e+15 9.62e-01 9.02e-01 2.59e+11 100 2.72e+01 6.51e+02
24 7.612014e+05 6.42e+01 5.17e+15 9.19e-01 9.21e-01 6.40e+11 100 2.92e+01 6.80e+02
25 7.611422e+05 5.92e+01 7.67e+15 9.54e-01 8.96e-01 1.28e+12 100 2.76e+01 7.08e+02
26 7.610808e+05 6.13e+01 1.02e+16 9.18e-01 9.05e-01 2.72e+12 100 2.73e+01 7.35e+02
27 7.610212e+05 5.96e+01 8.75e+15 9.42e-01 8.89e-01 5.14e+12 100 2.78e+01 7.63e+02
28 7.609591e+05 6.21e+01 9.05e+15 9.16e-01 9.05e-01 1.09e+13 100 2.84e+01 7.92e+02
29 7.609035e+05 5.56e+01 8.75e+15 9.29e-01 8.42e-01 1.61e+13 100 2.85e+01 8.20e+02
30 7.608408e+05 6.27e+01 1.19e+16 9.11e-01 8.78e-01 2.84e+13 100 2.93e+01 8.49e+02
31 7.607817e+05 5.91e+01 9.08e+15 9.13e-01 8.59e-01 4.50e+13 100 3.04e+01 8.80e+02
32 7.607202e+05 6.15e+01 8.37e+15 9.06e-01 8.68e-01 7.48e+13 100 2.69e+01 9.07e+02
33 7.606560e+05 6.42e+01 7.75e+15 8.97e-01 9.09e-01 1.65e+14 100 3.12e+01 9.38e+02
34 7.605954e+05 6.05e+01 3.89e+15 8.97e-01 8.91e-01 3.16e+14 100 2.84e+01 9.66e+02
35 7.605378e+05 5.76e+01 2.41e+16 8.76e-01 8.48e-01 4.77e+14 100 2.81e+01 9.94e+02
36 7.604731e+05 6.48e+01 8.78e+15 8.87e-01 8.90e-01 9.11e+14 100 2.60e+01 1.02e+03
37 7.604132e+05 5.99e+01 9.21e+15 8.54e-01 8.70e-01 1.53e+15 100 2.72e+01 1.05e+03
38 7.604110e+05 2.24e+00 1.56e+16 8.11e-03 2.48e-01 1.36e+15 3 3.50e+00 1.05e+03
39 7.603444e+05 6.66e+01 4.66e+15 8.88e-01 9.40e-01 4.07e+15 100 2.66e+01 1.08e+03
40 7.602978e+05 4.66e+01 3.08e+15 9.62e-01 8.65e-01 6.66e+15 100 2.71e+01 1.10e+03
41 7.601076e+05 1.90e+02 1.36e+16 1.98e+00 9.67e-01 1.00e+16 100 2.76e+01 1.13e+03
42 7.599268e+05 1.81e+02 5.24e+15 1.33e+00 9.55e-01 1.00e+16 96 2.66e+01 1.16e+03
43 7.599262e+05 6.29e-01 1.76e+16 1.56e-02 7.87e-02 6.26e+15 4 4.53e+00 1.16e+03
44 7.599240e+05 2.15e+00 5.04e+15 2.65e-03 2.68e-01 5.69e+15 3 4.73e+00 1.17e+03
45 7.599267e+05 -2.66e+00 0.00e+00 1.73e-03 -5.01e-01 2.84e+15 3 2.78e+00 1.17e+03
46 7.599264e+05 -2.34e+00 0.00e+00 2.24e-03 -4.38e-01 7.11e+14 3 2.71e+00 1.17e+03
47 7.599253e+05 -1.26e+00 0.00e+00 2.26e-03 -2.36e-01 8.89e+13 3 3.26e+00 1.18e+03
48 7.599258e+05 -1.71e+00 0.00e+00 2.26e-03 -3.19e-01 5.56e+12 3 2.12e+00 1.18e+03
49 7.599257e+05 -1.65e+00 0.00e+00 2.26e-03 -3.09e-01 1.74e+11 3 2.94e+00 1.18e+03
50 7.599257e+05 -1.61e+00 0.00e+00 2.26e-03 -3.01e-01 2.71e+09 3 3.07e+00 1.19e+03
Bundle adjustment report
------------------------
Residuals : 4890898
Parameters : 849200
Iterations : 51
Time : 1185.66 [s]
Initial cost : 0.397433 [px]
Final cost : 0.394177 [px]
Termination : No convergence
=> Completed observations: 2256
=> Merged observations: 1141
=> Filtered observations: 574
=> Changed observations: 0.001624
==============================================================================
Global bundle adjustment
==============================================================================
iter cost cost_change |gradient| |step| tr_ratio tr_radius ls_iter iter_time total_time
0 7.619420e+05 0.00e+00 1.02e+16 0.00e+00 0.00e+00 1.00e+04 0 2.18e+00 1.50e+01
1 7.561923e+05 5.75e+03 1.80e+16 1.68e+00 1.00e+00 3.00e+04 34 1.41e+01 2.91e+01
2 7.555906e+05 6.02e+02 5.15e+15 1.77e+00 9.89e-01 9.00e+04 100 2.86e+01 5.77e+01
3 7.553218e+05 2.69e+02 6.10e+15 1.64e+00 9.79e-01 2.70e+05 100 2.71e+01 8.48e+01
4 7.551962e+05 1.26e+02 8.61e+15 1.44e+00 9.45e-01 8.10e+05 100 3.06e+01 1.15e+02
5 7.551334e+05 6.28e+01 9.50e+15 7.95e-01 8.83e-01 1.47e+06 100 3.01e+01 1.46e+02
6 7.550757e+05 5.77e+01 5.05e+15 6.89e-01 9.12e-01 3.33e+06 100 2.66e+01 1.72e+02
7 7.550260e+05 4.97e+01 1.37e+16 6.23e-01 8.68e-01 5.53e+06 100 2.78e+01 2.00e+02
8 7.549790e+05 4.70e+01 1.32e+16 5.62e-01 8.34e-01 7.86e+06 100 2.89e+01 2.29e+02
9 7.549288e+05 5.01e+01 1.15e+16 6.02e-01 8.80e-01 1.40e+07 100 2.81e+01 2.57e+02
10 7.548792e+05 4.97e+01 3.63e+15 5.41e-01 9.13e-01 3.21e+07 100 2.69e+01 2.84e+02
11 7.548343e+05 4.49e+01 5.60e+15 6.06e-01 8.61e-01 5.17e+07 100 2.83e+01 3.12e+02
12 7.547897e+05 4.46e+01 2.49e+16 5.35e-01 8.38e-01 7.48e+07 100 2.84e+01 3.41e+02
13 7.547427e+05 4.70e+01 1.52e+16 6.06e-01 8.63e-01 1.21e+08 100 2.83e+01 3.69e+02
14 7.546961e+05 4.66e+01 8.15e+15 5.38e-01 8.89e-01 2.29e+08 100 2.90e+01 3.98e+02
15 7.546507e+05 4.55e+01 1.06e+16 6.01e-01 8.98e-01 4.60e+08 100 2.83e+01 4.26e+02
16 7.546089e+05 4.18e+01 8.61e+15 5.40e-01 8.37e-01 6.64e+08 100 2.77e+01 4.54e+02
17 7.545675e+05 4.14e+01 1.68e+16 5.94e-01 8.19e-01 8.98e+08 100 2.67e+01 4.81e+02
18 7.545184e+05 4.90e+01 4.54e+15 5.43e-01 9.34e-01 2.58e+09 100 2.98e+01 5.10e+02
19 7.544803e+05 3.81e+01 1.78e+16 5.87e-01 8.16e-01 3.45e+09 100 2.94e+01 5.40e+02
20 7.544360e+05 4.43e+01 9.00e+15 5.45e-01 8.67e-01 5.72e+09 100 2.81e+01 5.68e+02
21 7.543977e+05 3.84e+01 1.33e+16 5.79e-01 7.97e-01 7.24e+09 100 2.60e+01 5.94e+02
22 7.543534e+05 4.43e+01 1.58e+16 5.48e-01 8.56e-01 1.13e+10 100 2.73e+01 6.21e+02
23 7.543136e+05 3.98e+01 1.44e+16 5.72e-01 8.24e-01 1.55e+10 100 2.56e+01 6.47e+02
24 7.542709e+05 4.27e+01 1.75e+16 5.52e-01 8.44e-01 2.30e+10 100 2.87e+01 6.76e+02
25 7.542277e+05 4.32e+01 1.19e+16 5.70e-01 8.88e-01 4.32e+10 100 2.80e+01 7.04e+02
26 7.541887e+05 3.90e+01 1.29e+16 5.60e-01 8.20e-01 5.86e+10 100 2.69e+01 7.30e+02
27 7.541487e+05 4.01e+01 9.83e+15 5.68e-01 8.21e-01 7.96e+10 100 2.70e+01 7.57e+02
28 7.541037e+05 4.49e+01 7.86e+15 5.69e-01 8.77e-01 1.39e+11 100 2.85e+01 7.86e+02
29 7.540626e+05 4.11e+01 5.56e+15 5.68e-01 8.72e-01 2.37e+11 100 2.87e+01 8.15e+02
30 7.540224e+05 4.02e+01 2.23e+16 5.80e-01 8.50e-01 3.61e+11 100 2.71e+01 8.42e+02
31 7.539843e+05 3.81e+01 2.52e+16 5.71e-01 7.89e-01 4.47e+11 100 2.80e+01 8.70e+02
32 7.539368e+05 4.75e+01 2.94e+15 5.93e-01 9.14e-01 1.04e+12 100 2.61e+01 8.96e+02
33 7.539013e+05 3.55e+01 1.34e+16 5.77e-01 7.83e-01 1.27e+12 100 2.87e+01 9.24e+02
34 7.538597e+05 4.15e+01 1.61e+16 6.06e-01 8.12e-01 1.67e+12 100 2.83e+01 9.53e+02
35 7.538156e+05 4.41e+01 8.01e+15 5.83e-01 8.68e-01 2.78e+12 100 2.80e+01 9.81e+02
36 7.537764e+05 3.92e+01 6.49e+15 6.22e-01 8.25e-01 3.83e+12 100 2.72e+01 1.01e+03
37 7.537352e+05 4.12e+01 1.76e+16 5.93e-01 8.30e-01 5.37e+12 100 2.74e+01 1.04e+03
38 7.536936e+05 4.16e+01 6.07e+15 6.39e-01 8.41e-01 7.86e+12 100 2.67e+01 1.06e+03
39 7.536494e+05 4.42e+01 4.65e+15 6.03e-01 9.02e-01 1.64e+13 100 2.94e+01 1.09e+03
40 7.536120e+05 3.75e+01 4.98e+15 6.58e-01 8.06e-01 2.13e+13 100 2.90e+01 1.12e+03
41 7.535699e+05 4.21e+01 1.22e+16 6.18e-01 8.71e-01 3.59e+13 100 2.70e+01 1.15e+03
42 7.535296e+05 4.03e+01 1.85e+16 6.80e-01 8.40e-01 5.25e+13 100 2.65e+01 1.17e+03
43 7.534878e+05 4.18e+01 6.25e+15 6.35e-01 8.68e-01 8.75e+13 100 2.89e+01 1.20e+03
44 7.534472e+05 4.06e+01 1.10e+16 7.01e-01 8.55e-01 1.37e+14 100 2.69e+01 1.23e+03
45 7.534066e+05 4.06e+01 7.07e+15 6.53e-01 8.49e-01 2.06e+14 100 3.06e+01 1.26e+03
46 7.533641e+05 4.26e+01 3.75e+15 7.25e-01 8.91e-01 3.97e+14 100 2.87e+01 1.29e+03
47 7.533227e+05 4.14e+01 2.43e+15 6.74e-01 9.03e-01 8.30e+14 100 2.53e+01 1.31e+03
48 7.532949e+05 2.77e+01 9.98e+15 6.72e-01 7.94e-01 1.04e+15 100 2.68e+01 1.34e+03
49 7.532401e+05 5.48e+01 2.54e+16 7.79e-01 8.25e-01 1.44e+15 100 2.75e+01 1.37e+03
50 7.531789e+05 6.12e+01 3.52e+15 6.41e-01 9.34e-01 4.13e+15 100 2.87e+01 1.40e+03
Bundle adjustment report
------------------------
Residuals : 4894008
Parameters : 848930
Iterations : 51
Time : 1397.96 [s]
Initial cost : 0.394574 [px]
Final cost : 0.392299 [px]
Termination : No convergence
=> Completed observations: 870
=> Merged observations: 879
=> Filtered observations: 458
=> Changed observations: 0.000902
=> Filtered images: 0
Elapsed time: 299.162 [minutes]
最开始 kitti 04 的 triangulation 情况：
nohup: ignoring input
==============================================================================
Loading model
==============================================================================
==============================================================================
Loading database
==============================================================================
Loading cameras... 534 in 0.001s
Loading matches... 22761 in 0.128s
Loading images... 271 in 0.194s (connected 271)
Building correspondence graph... in 0.558s (ignored 0)
Elapsed time: 0.015 [minutes]
==============================================================================
Triangulating image #4 (0)
==============================================================================
=> Image sees 0 / 2804 points
=> Triangulated 84 points
==============================================================================
Triangulating image #7 (1)
==============================================================================
=> Image sees 46 / 3346 points
=> Triangulated 73 points
==============================================================================
Triangulating image #8 (2)
==============================================================================
=> Image sees 57 / 3403 points
=> Triangulated 75 points
==============================================================================
Triangulating image #1 (3)
==============================================================================
=> Image sees 71 / 3271 points
=> Triangulated 51 points
==============================================================================
Triangulating image #22 (4)
==============================================================================
=> Image sees 79 / 3334 points
=> Triangulated 82 points
==============================================================================
Triangulating image #16 (5)
==============================================================================
=> Image sees 100 / 3308 points
=> Triangulated 78 points
==============================================================================
Triangulating image #10 (6)
==============================================================================
=> Image sees 99 / 3394 points
=> Triangulated 73 points
==============================================================================
Triangulating image #11 (7)
==============================================================================
=> Image sees 107 / 3481 points
=> Triangulated 75 points
==============================================================================
Triangulating image #2 (8)
==============================================================================
=> Image sees 101 / 3474 points
=> Triangulated 77 points
==============================================================================
Triangulating image #3 (9)
==============================================================================
=> Image sees 120 / 3464 points
=> Triangulated 74 points
==============================================================================
Triangulating image #5 (10)
==============================================================================
=> Image sees 111 / 3447 points
=> Triangulated 99 points
==============================================================================
Triangulating image #6 (11)
==============================================================================
=> Image sees 113 / 3479 points
=> Triangulated 76 points
==============================================================================
Triangulating image #17 (12)
==============================================================================
=> Image sees 126 / 3458 points
=> Triangulated 97 points
==============================================================================
Triangulating image #25 (13)
==============================================================================
=> Image sees 138 / 3385 points
=> Triangulated 113 points
==============================================================================
Triangulating image #28 (14)
==============================================================================
=> Image sees 140 / 3251 points
=> Triangulated 96 points
==============================================================================
Triangulating image #21 (15)
==============================================================================
=> Image sees 134 / 3180 points
=> Triangulated 94 points
==============================================================================
Triangulating image #27 (16)
==============================================================================
=> Image sees 141 / 3126 points
=> Triangulated 114 points
==============================================================================
Triangulating image #24 (17)
==============================================================================
=> Image sees 151 / 2969 points
=> Triangulated 113 points
==============================================================================
Triangulating image #23 (18)
==============================================================================
=> Image sees 192 / 2915 points
=> Triangulated 113 points
==============================================================================
Triangulating image #15 (19)
==============================================================================
=> Image sees 181 / 2820 points
=> Triangulated 108 points
==============================================================================
Triangulating image #13 (20)
==============================================================================
=> Image sees 179 / 2790 points
=> Triangulated 117 points
==============================================================================
Triangulating image #12 (21)
==============================================================================
=> Image sees 187 / 2797 points
=> Triangulated 117 points
==============================================================================
Triangulating image #9 (22)
==============================================================================
=> Image sees 176 / 2723 points
=> Triangulated 109 points
==============================================================================
Triangulating image #14 (23)
==============================================================================
=> Image sees 189 / 2746 points
=> Triangulated 129 points
==============================================================================
Triangulating image #26 (24)
==============================================================================
=> Image sees 197 / 2745 points
=> Triangulated 146 points
==============================================================================
Triangulating image #19 (25)
==============================================================================
=> Image sees 205 / 2706 points
=> Triangulated 126 points
==============================================================================
Triangulating image #20 (26)
==============================================================================
=> Image sees 190 / 2786 points
=> Triangulated 134 points
==============================================================================
Triangulating image #18 (27)
==============================================================================
=> Image sees 210 / 2754 points
=> Triangulated 139 points
==============================================================================
Triangulating image #35 (28)
==============================================================================
=> Image sees 213 / 2778 points
=> Triangulated 143 points
==============================================================================
Triangulating image #31 (29)
==============================================================================
=> Image sees 193 / 2733 points
=> Triangulated 151 points
==============================================================================
Triangulating image #38 (30)
==============================================================================
=> Image sees 177 / 2852 points
=> Triangulated 155 points
==============================================================================
Triangulating image #34 (31)
==============================================================================
=> Image sees 174 / 2807 points
=> Triangulated 161 points
==============================================================================
Triangulating image #36 (32)
==============================================================================
=> Image sees 179 / 2849 poin
最低0.47元/天 解锁文章
扫一扫
1934
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
