Too large memory required？

[whu-lyh]

Hi, koide, thanks for your works. I tried your registration method, specifically fast voxel boosted gicp, on a more dense point clouds. I found that the memory required is too big for correspondence searching, as the following code shows

void FastVGICP<PointSource, PointTarget>::update_correspondences(const Eigen::Isometry3d& trans) {
		voxel_correspondences_.clear();
		auto offsets = neighbor_offsets(search_method_);
		unsigned int scaled_pts_num = input_->size() * offsets.size();
		std::vector<std::vector<std::pair<int, GaussianVoxel::Ptr>>> corrs(num_threads_);
		for (auto& c : corrs) {
			c.reserve(scaled_pts_num / num_threads_);
		}
                ...
		voxel_correspondences_.reserve(scaled_pts_num);
                ...
}

or more details at

fast_gicp/include/fast_gicp/gicp/impl/fast_vgicp_impl.hpp

Line 96 in 2e9fd0b

voxel_correspondences_.reserve(input_->size() * offsets.size());

and at

fast_gicp/include/fast_gicp/gicp/impl/fast_vgicp_impl.hpp

Line 102 in 2e9fd0b

voxel_mahalanobis_.resize(voxel_correspondences_.size());

My question is if there are large amount of pts inside a voxel, the memory cost is too large for a simple 7nn neighbor. Although i know this operation is used for fast query neighborhood pts, but memory cost is still quite large? My laptop doesn't have enough memory for such a large memory cost operation.