search

AIBluefisher / GraphOptim

The official implementation of our CVPR 2021 paper - Hybrid Rotation Averaging: A Fast and Robust Rotation Averaging Approach
BSD 3-Clause "New" or "Revised" License
147 stars 24 forks source link

$VOC_TREE_PATH $MOST_SIMILAR_IMAGES_NUM #12

Open theyunfeng opened 1 year ago

theyunfeng commented 1 year ago

你好，小白问一下这两个参数分别指的是什么？第一个时词典路径还是指什么？需要是什么格式呢？后一个参数是随便设置吗？

AIBluefisher commented 1 year ago

$VOC_TREE_PATH denotes the absolute path of the vocabulary tree provided by COLMAP. The second parameter denotes the upper bound of the images you want to match (you need to adjust it according to the scale of your own dataset).

theyunfeng commented 1 year ago

image 这里是不是对应的应该是points2才对...

AIBluefisher commented 1 year ago

you're right. Which file is it? I'll fix it soon.

theyunfeng commented 1 year ago

you're right. Which file is it? I'll fix it soon.

global_mapper.cpp里面的函数RefineRelativeTranslationWithKnownRotation(...)

theyunfeng commented 1 year ago

you're right. Which file is it? I'll fix it soon.

顺便请问一下，这个错误是什么原因的？看着像是求解过程为之原因崩溃了，看样子求解器好像是您自己写的，所以想问一下您有没有头绪，莫非是我自己的数据集重复纹理过多或其他噪音太多导致数据较差而导致求解器无法求解？ image

AIBluefisher commented 1 year ago

It may be due to the view graph being too sparse and it contains multiple connected components. It can try to identify if there exists more than one connected component in the view graph. If so, you can try to increase the number of candidate matching pairs for each view to obtain a denser graph.