Semantic Labeling of Mobile LiDAR Point Clouds via Active Learning and Higher Order MRF

Luo, Haipeng; Wang, Cheng; Wen, Chenglu; Chen, Ziyi; Zai, Dawei; Yu, Yongtao; Li, Jonathan

doi:10.1109/tgrs.2018.2802935

Cited by 57 publications

(39 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Forest in situ measurements by non-professional users, such as forest owners, has become feasible by using, e.g., a cellphone camera. Moreover, in recent years, mobile and personal mapping have become available (Liang et al 2014;Bauwens et al 2016;Forsman et al 2016b;Marselis et al 2016;Juraj et al 2017;Oveland et al 2017) and many efforts have been invested to new Mobile Laser Scanning (MLS)-relevant data processing Liang et al 2018b;Luo et al 2018). Mobile and personal mapping systems integrating the LS/camera sensors, the kinematic platforms, and/or the navigation sensors, are capable to measure the forest plots 3-20 times faster than stationary systems (Liang et al 2014).…”

Section: Introductionmentioning

confidence: 99%

Forest in situ observations using unmanned aerial vehicle as an alternative of terrestrial measurements

et al. 2019

View full text Add to dashboard Cite

Background: Lately, terrestrial point clouds have drawn attention as a new data source for in situ forest investigations. So far, terrestrial laser scanning (TLS) has the highest data quality among all terrestrial point cloud data in terms of geometric accuracy and level of detail (IEEE Transact Geosci Remote Sens 53: 5117-5132, 2015). The TLS point clouds processed by automated algorithms can provide certain individual tree parameters at close to required accuracy in practical applications. However, all terrestrial point clouds face a general challenge, which is the occlusions of upper tree crowns. An emerging technology called unmanned-aerial-vehicle (UAV) -borne laser scanning (ULS) potentially combines the strengths of above and under canopy surveys.Results: The performance of ULS are evaluated in 22 sample plots of various forest stand conditions in a boreal forest. The forest parameter estimates are benchmarked through a comparison with state-of-the-art terrestrial mechanisms from both static terrestrial and mobile laser scanning. The results show that in easy forest stand conditions, the performance of ULS point cloud is comparable with the terrestrial solutions. Conclusions: This study gives the first strict evaluation of ULS in situ observations in varied forest conditions. The study also acts as a benchmarking of available active remote sensing techniques for forest in situ mensuration. The results indicate that the current off-the-shelf ULS has an excellent tree height/tops measurement performance. Although the geometrical accuracy of the ULS data, especially at the stem parts, does not yet reach the level of other terrestrial point clouds, the unbeatable high mobility and fast data acquisition make the ULS a very attractive option in forest investigations.

show abstract

Section: Introductionmentioning

confidence: 99%

Forest in situ observations using unmanned aerial vehicle as an alternative of terrestrial measurements

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Zhu et al [3] proposed the generation of supervoxels to merge neighborhoods according to their features. Recently, a similar strategy using supervoxels was discussed by Lou et al [13]. In addition to using a single primitive to represent a scanning scene, researchers also investigated how to exploit different primitives' representations.…”

Section: Related Workmentioning

confidence: 99%

“…Due to its excellent performance, the RF classifier has received much attention [18]. In a recent study, many researchers [2,4,9,13,15] implemented the RF classifier to carry out the supervised classification. To obtain spatially smoothing semantic labels of 3D point clouds in multi-class cases, an approximate optimal strategy can be applied using the initial classification and contextual information [2,3].…”

Section: Related Workmentioning

confidence: 99%

“…Contextual information has played a very important role in the recent classification of point clouds [2,3,13,29]. According to the different discriminative powers of different primitives, we categorize the geometric relationships into two grades.…”

Section: Pairwise Featuresmentioning

confidence: 99%

“…So far, there is no unified definition of high-order features [2]. In many applications, the regional context (namely the long-range relationships and label cost) is used to calculate high-order potentials [13,30,31]. Although such a definition makes sense for the classification of point clouds if the unary potentials are based on the point primitive or the (super) voxel primitive, it is not the best definition in cases concerning the segment primitive.…”

Section: Heuristic Rule-based Featuresmentioning

confidence: 99%

See 2 more Smart Citations

A Multi-Primitive-Based Hierarchical Optimal Approach for Semantic Labeling of ALS Point Clouds

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

There are normally three main steps to carrying out the labeling of airborne laser scanning (ALS) point clouds. The first step is to use appropriate primitives to represent the scanning scenes, the second is to calculate the discriminative features of each primitive, and the third is to introduce a classifier to label the point clouds. This paper investigates multiple primitives to effectively represent scenes and exploit their geometric relationships. Relationships are graded according to the properties of related primitives. Then, based on initial labeling results, a novel, hierarchical, and optimal strategy is developed to optimize semantic labeling results. The proposed approach was tested using two sets of representative ALS point clouds, namely the Vaihingen datasets and Hong Kong’s Central District dataset. The results were compared with those generated by other typical methods in previous work. Quantitative assessments for the two experimental datasets showed that the performance of the proposed approach was superior to reference methods in both datasets. The scores for correctness attained over 98% in all cases of the Vaihingen datasets and up to 96% in the Hong Kong dataset. The results reveal that our approach of labeling different classes in terms of ALS point clouds is robust and bears significance for future applications, such as 3D modeling and change detection from point clouds.

show abstract

LiDAL: Inter-frame Uncertainty Based Active Learning for 3D LiDAR Semantic Segmentation

Bai

Wang

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

We propose LiDAL, a novel active learning method for 3D Li-DAR semantic segmentation by exploiting inter-frame uncertainty among LiDAR frames. Our core idea is that a well-trained model should generate robust results irrespective of viewpoints for scene scanning and thus the inconsistencies in model predictions across frames provide a very reliable measure of uncertainty for active sample selection. To implement this uncertainty measure, we introduce new inter-frame divergence and entropy formulations, which serve as the metrics for active selection. Moreover, we demonstrate additional performance gains by predicting and incorporating pseudo-labels, which are also selected using the proposed inter-frame uncertainty measure. Experimental results validate the effectiveness of LiDAL: we achieve 95% of the performance of fully supervised learning with less than 5% of annotations on the SemanticKITTI and nuScenes datasets, outperforming state-of-the-art active learning methods. Code release: https://github.com/hzykent/LiDAL

show abstract

Semantic Labeling of Mobile LiDAR Point Clouds via Active Learning and Higher Order MRF

Cited by 57 publications

References 33 publications

Forest in situ observations using unmanned aerial vehicle as an alternative of terrestrial measurements

Forest in situ observations using unmanned aerial vehicle as an alternative of terrestrial measurements

A Multi-Primitive-Based Hierarchical Optimal Approach for Semantic Labeling of ALS Point Clouds

LiDAL: Inter-frame Uncertainty Based Active Learning for 3D LiDAR Semantic Segmentation

Contact Info

Product

Resources

About