Consistent Mapping Research Articles

State‐contingent production technology formulation: Identifying states of nature using reduced‐form econometric models of crop yield

AbstractConducting experiments can be time consuming and expensive, and may not always be reasonable. Therefore, empirical research often derives structural parameters based on observational data and reduced‐form econometric models. The state‐contingent approach presents a consistent conceptual framework for analyzing producer decisions under uncertainty. However, application of this structural modeling approach has been hampered by data constraints, particularly the lack of information for mapping producers' stochastic outputs onto a set of the states of nature representing different uncertain events. Consistent mapping of uncertainty is particularly critical in the context of multiple output production where weather shocks often have different effects across crops and in microeconometric analyses when unobserved farm heterogeneity may confound the effect of uncertainty. Our study demonstrates how the application of reduced‐form approaches can overcome constraints of structural econometric modeling associated with the lack of relevant data and presents an approach for identifying states of nature in the context of multiple output production using reduced‐form econometric models of crop yield. In an empirical application based on Hungarian farm accountancy data, we demonstrate that the proposed approach allows a consistent mapping of production uncertainty in crop farming, utilizes panel data structure, and controls for potential endogeneity due to unobserved farm heterogeneity. We anticipate the presented approach to be useful for developing further the state‐contingent approach and to stimulate further studies combining the strengths of structural approaches and reduced‐form models.

Metric networks for enhanced perception of non-local semantic information.

Metric learning, as a fundamental research direction in the field of computer vision, has played a crucial role in image matching. Traditional metric learning methods aim at constructing two-branch siamese neural networks to address the challenge of image matching, but they often overlook to cross-source and cross-view scenarios. In this article, a multi-branch metric learning model is proposed to address these limitations. The main contributions of this work are as follows: Firstly, we design a multi-branch siamese network model that enhances measurement reliability through information compensation among data points. Secondly, we construct a non-local information perception and fusion model, which accurately distinguishes positive and negative samples by fusing information at different scales. Thirdly, we enhance the model by integrating semantic information and establish an information consistency mapping between multiple branches, thereby improving the robustness in cross-source and cross-view scenarios. Experimental tests which demonstrate the effectiveness of the proposed method are carried out under various conditions, including homologous, heterogeneous, multi-view, and crossview scenarios. Compared to the state-of-the-art comparison algorithms, our proposed algorithm achieves an improvement of ~1, 2, 1, and 1% in terms of similarity measurement Recall@10, respectively, under these four conditions. In addition, our work provides an idea for improving the crossscene application ability of UAV positioning and navigation algorithm.

Phenology-assisted supervised paddy rice mapping with the Landsat imagery on Google Earth Engine: Experiments in Heilongjiang Province of China from 1990 to 2020

Accurate spatial distribution maps of paddy rice played crucial roles in food security and market stability. Decades-spanning Landsat images were useful for long-term paddy rice mapping. However, it still remained challenging to achieve consistent paddy rice mapping using the Landsat series images due to many factors such as sparse observations, frequent weather contamination, and shortage of training samples. To address these challenges, this study proposed a flexible Phenology-assisted Supervised Paddy Rice (PSPR) mapping framework on Google Earth Engine (GEE). This was achieved by utilizing the automation of the phenological methods, training data generation and purification, and the all-season classification capacity of the machine learning methods. We demonstrated the method by generating high-resolution 30-m paddy rice maps of Heilongjiang Province of China from 1990 to 2020. The derived rice maps were validated using abundant reference samples, four existing paddy rice products, and available agricultural statistics. The result showed that improved performance was verified in comparison to previous studies and a high linear relationship was observed with an average R2 of 0.993. Based on the spatiotemporal analysis, it was discovered that the rice planting in Heilongjiang has significantly shifted northward in the last three decades and this northward shift surprisingly appeared earlier than the previous studies, which was to our best knowledge first to be revealed in related studies. The multi-year dataset is useful for rice monitoring, water management, and policy making. All data and codes used in this study can be accessed on GitHub (https://github.com/MKGenesis/PSPR-rice-HLJ).

A multidimensional characterization of the neurocognitive architecture underlying age-related temporal speech processing

Healthy aging is often associated with speech comprehension difficulties in everyday life situations despite a pure-tone hearing threshold in the normative range. Drawing on this background, we used a multidimensional approach to assess the functional and structural neural correlates underlying age-related temporal speech processing while controlling for pure-tone hearing acuity. Accordingly, we combined structural magnetic resonance imaging and electroencephalography, and collected behavioral data while younger and older adults completed a phonetic categorization and discrimination task with consonant-vowel syllables varying along a voice-onset time continuum. The behavioral results confirmed age-related temporal speech processing singularities which were reflected in a shift of the boundary of the psychometric categorization function, with older adults perceiving more syllable characterized by a short voice-onset time as /ta/ compared to younger adults. Furthermore, despite the absence of any between-group differences in phonetic discrimination abilities, older adults demonstrated longer N100/P200 latencies as well as increased P200 amplitudes while processing the consonant-vowel syllables varying in voice-onset time. Finally, older adults also exhibited a divergent anatomical gray matter infrastructure in bilateral auditory-related and frontal brain regions, as manifested in reduced cortical thickness and surface area. Notably, in the younger adults but not in the older adult cohort, cortical surface area in these two gross anatomical clusters correlated with the categorization of consonant-vowel syllables characterized by a short voice-onset time, suggesting the existence of a critical gray matter threshold that is crucial for consistent mapping of phonetic categories varying along the temporal dimension. Taken together, our results highlight the multifaceted dimensions of age-related temporal speech processing characteristics, and pave the way toward a better understanding of the relationships between hearing, speech and the brain in older age.

Reduction in sugar drink valuation and consumption with gamified executive control training

The overvaluation of high-energy, palatable food cues contributes to unhealthy eating and being overweight. Reducing the valuation of unhealthy food may thus constitute a powerful lever to improve eating habits and conditions characterized by unhealthy eating. We conducted a double-blind, placebo-controlled, randomized intervention trial assessing the efficacy of a five to twenty days online cognitive training intervention to reduce sugary drink perceived palatability and consumption. Our intervention involved a recently identified action-to-valuation mechanism of action, in which the repeated inhibition of prepotent motor responses to hedonic food cues in a Go/NoGo (GNG) and an attentional bias modification (ABM) task eventually reduces their valuation and intake. Confirming our hypotheses, the experimental intervention with consistent (100%) mapping between motor inhibition and the targeted unhealthy sugary drinks cues induced a larger decrease in their valuation than the control intervention with inconsistent (50%) mapping (− 27.6% vs. − 19%), and a larger increase of the (water) items associated with response execution (+ 11% vs + 4.2%). Exploratory analyses suggest that the effect of training on unhealthy items valuation may persist for at least one month. Against our hypothesis, we observed equivalent reductions in self-reported consumption of sugary drinks following the two interventions (exp: − 27% vs. ctrl: − 19%, BF01 = 4.7), suggesting a dose-independent effect of motor inhibition on self-reported consumption. Our collective results corroborate the robustness and large size of the devaluation effects induced by response inhibition on palatable items, but challenge the assumption of a linear relationship between such effects and the actual consumption of the target items.Protocol registrationThe stage 1 protocol for this Registered Report was accepted in principle on 30/03/2021. The protocol, as accepted by the journal, can be found at: https://doi.org/10.17605/OSF.IO/5ESMP.

Exploring crossmodal correspondences for future research in human movement augmentation.

"Crossmodal correspondences" are the consistent mappings between perceptual dimensions or stimuli from different sensory domains, which have been widely observed in the general population and investigated by experimental psychologists in recent years. At the same time, the emerging field of human movement augmentation (i.e., the enhancement of an individual's motor abilities by means of artificial devices) has been struggling with the question of how to relay supplementary information concerning the state of the artificial device and its interaction with the environment to the user, which may help the latter to control the device more effectively. To date, this challenge has not been explicitly addressed by capitalizing on our emerging knowledge concerning crossmodal correspondences, despite these being tightly related to multisensory integration. In this perspective paper, we introduce some of the latest research findings on the crossmodal correspondences and their potential role in human augmentation. We then consider three ways in which the former might impact the latter, and the feasibility of this process. First, crossmodal correspondences, given the documented effect on attentional processing, might facilitate the integration of device status information (e.g., concerning position) coming from different sensory modalities (e.g., haptic and visual), thus increasing their usefulness for motor control and embodiment. Second, by capitalizing on their widespread and seemingly spontaneous nature, crossmodal correspondences might be exploited to reduce the cognitive burden caused by additional sensory inputs and the time required for the human brain to adapt the representation of the body to the presence of the artificial device. Third, to accomplish the first two points, the benefits of crossmodal correspondences should be maintained even after sensory substitution, a strategy commonly used when implementing supplementary feedback.

Spelling-sound consistency influences second-language age of acquisition effect: evidence for the arbitrary mapping hypothesis

ABSTRACT The age at which people acquire a word influences word recognition, known as the age of acquisition (AoA) effect. In the first language (L1), AoA effects are widely found in various languages and experimental tasks. Arbitrary Mapping Hypothesis proposes that AoA effects reflect the loss of network plasticity during the learning of mappings between input and output representations. It predicts that the AoA effect appears (or is larger) when the input-output mapping is arbitrary/inconsistent, relative to consistent input-output mapping. The present study examined how these predictions generalised to the second language (L2). We explored whether the L2 AoA effect was modulated by spelling-sound consistency in Chinese-English bilinguals, adopting a delayed word naming task. The results showed that the mapping consistency modulated the L2 AoA effect on the N170, P200 and N400 components. L2 AoA effect was insignificant in the consistent condition but was significant in the inconsistent condition: late-acquired words elicited larger N170, smaller P200, and larger N400 compared to early-acquired words. These findings suggest that L2 AoA effects occur in the spelling-sound connections, providing direct evidence for the Arbitrary Mapping Hypothesis.

MODE: Monocular omnidirectional depth estimation via consistent depth fusion

Monocular depth estimation has seen significant progress in recent years, especially in outdoor scenes. However, depth estimation results are not satisfying in omnidirectional images. As compared to perspective images, estimating the depth map from an omnidirectional image captured in the outdoor scene, using neural networks, has two additional challenges: (i) the depth range of outdoor images varies a lot across different scenes, making it difficult for the depth network to predict accurate depth results for training with an indoor dataset, besides the maximum distance in outdoor scenes mostly stays the same as the camera sees the sky, but depth labels in this region are entirely missing in existing datasets; (ii) a standard representation of omnidirectional images introduces spherical distortion, which causes difficulties for the vanilla network to predict accurate relative structural depth details. In this paper, we propose a novel network-MODE by giving special considerations to those challenges and designing a set of flexible modules for improving the performance of omnidirectional depth estimation. First, a consistent depth structure module is proposed to estimate a consistent depth structure map, and the predicted structural map can improve depth details. Second, to suit the characteristics of spherical sampling, we propose a strip convolution fusion module to enhance long-range dependencies. Third, rather than using a single depth decoder branch as in previous methods, we propose a semantics decoder branch to estimate sky regions in the omnidirectional image. The proposed method is validated on three widely used datasets, demonstrating the state-of-the-art performance. Moreover, the effectiveness of each module is shown through an ablation study on real-world datasets. Our code is available at https://github.com/lkku1/MODE.© 2017 Elsevier Inc. All rights reserved.

The acquisition of English modal constructions: a corpus-based analysis.

The English modal system is complex, exhibitingmany-to-one, and one-to-many, form-function mappings. Usage-based approaches emphasise the role of the input in acquisition but rarely address the impact of form-function mappings on acquisition. To test whether consistent form-function mappings facilitate acquisition,we analysed two dense mother-child corpora at age 3 and 4. We examined the influence on acquisition of input features including form-function mapping frequency and the number of functions a modal signifies, using innovative methodological controlsfor other aspects of the input (e.g., form frequency) and child characteristics (e.g., age as a proxy for socio-cognitive development). The children weremore likely to produce the frequent modals and form-function mappings of their input but modals with fewer functions in caregiver speech did not promote acquisition of these forms. Our findings support usage-based approaches to language acquisition and demonstrate the importance of applying appropriatecontrols when investigating relationships between input and development.

Robust Loop Closure Selection Based on Inter-Robot and Intra-Robot Consistency for Multi-Robot Map Fusion

In multi-robot simultaneous localization and mapping (SLAM) systems, the system must create a consistent global map with multiple local maps and loop closures between robot poses. However, false-positive loop closures caused by perceptual aliasing can severely distort the global map, especially in GNSS-denied areas, where a good prior of relative poses between robots is unavailable. In addition, the performance of the consistency metric in existing map fusion methods relies on accurate odometry from each robot. However, in practice, cumulative noise is inevitably present in robot trajectories, which leads to poor map fusion with existing methods. Thus, in this paper, we propose a robust consistency-based inter-robot and intra-robot loop closure selection algorithm for map fusion. We consider both pairwise-loop consistency and loop-odometry consistency to improve robustness against false-positive loop closures and accumulative noise in the odometry. Specifically, we select a reliable inter-robot loop closure measurement with a consistency-based strategy to provide an initial prior of relative pose between two robot trajectories and update the pose variables of the robot trajectories. The loop closure selection problem is formulated as a maximum edge weight clique problem in graph theory. A performance evaluation of the proposed method was conducted on the ManhattanOlson3500, modified CSAIL and Bicocca datasets, and the experimental results demonstrate that the proposed method outperforms the pairwise consistency measurement set maximization method (PCM) under severe accumulative noise and can be integrated with M-estimation methods.

Environmental change and ecosystem functioning drive transitions in social-ecological systems: A stylized modelling approach

Sustainable management of social-ecological systems requires an understanding of how anthropogenic climate- and land use change may disrupt interactions between human societies and the ecosystem processes they depend on. In this study, we expand an existing stylized social-ecological system model by explicitly considering how urbanizing societies may become less dependent on local ecosystem functioning. This expansion is motivated by a previously developed conceptual framework suggesting that societies may reside in either a green loop and be strongly dependent on local ecosystem processes, or in a red loop where this dependency is weaker due to imports of natural resources from elsewhere. Analyzing the feasibility and stability of local social-ecological system states over a wide range of environmental and socio-economic conditions, we observed dynamics consistent with the notion of green loop-dominated and red loop-dominated societies comprising alternate stable social-ecological states. Based on systems' inherent dependencies on local ecosystem processes, responses to environmental change could comprise either transitions between green loop- and red loop-dominated states, or collapse of either of these states. Our quantitative model provides an internally consistent mapping of green loop- and red loop-dominated states, as well as transitions between or collapses of these states, along a gradient of environmental conditions.

On Farmland and Floodplains—Modeling Urban Growth Impacts Based on Global Population Scenarios in Pune, India

Emerging megacities in the global south face unprecedented transformation dynamics, manifested in rapid demographic, economic, and physical growth. Anticipating the associated sustainability and resilience challenges requires an understanding of future trajectories. Global change models provide consistent high-level urbanization scenarios. City-scale urban growth models accurately simulate complex physical growth. Modeling approaches linking the global and the local scale, however, are underdeveloped. This work introduces a novel approach to inform a local urban growth model by global Shared Socioeconomic Pathways to produce consistent maps of future urban expansion and population density via cellular automaton and dasymetric mapping. We demonstrate the approach for the case of Pune, India. Three scenarios are explored until 2050: business as usual (BAU), high, and low urbanization. After calibration and validation, the BAU scenario yields a 55% growth in Pune’s population and 90% in built-up extent, entailing significant impacts: Pune’s core city densifies further with up to 60,000 persons/km2, adding pressure to its strained infrastructure. In addition, 66–70% more residents are exposed to flood risk. Half of the urban expansion replaces agriculture, converting 167 km2 of land. The high-urbanization scenario intensifies these impacts. These results illustrate how spatially explicit scenario projections help identify impacts of urbanization and inform long-term planning.

Thickness Mapping and Layer Number Identification of Exfoliated van der Waals Materials by Fourier Imaging Micro-Ellipsometry

As performance of van der Waals heterostructure devices is governed by the nanoscale thicknesses and homogeneity of their constituent mono- to few-layer flakes, accurate mapping of these properties with high lateral resolution becomes imperative. Spectroscopic ellipsometry is a promising optical technique for such atomically thin-film characterization due to its simplicity, noninvasive nature and high accuracy. However, the effective use of standard ellipsometry methods on exfoliated micron-scale flakes is inhibited by their tens-of-microns lateral resolution or slow data acquisition. In this work, we demonstrate a Fourier imaging spectroscopic micro-ellipsometry method with sub-5 μm lateral resolution and three orders-of-magnitude faster data acquisition than similar-resolution ellipsometers. Simultaneous recording of spectroscopic ellipsometry information at multiple angles results in a highly sensitive system, which is used for performing angstrom-level accurate and consistent thickness mapping on exfoliated mono-, bi- and trilayers of graphene, hexagonal boron nitride (hBN) and transition metal dichalcogenide (MoS2, WS2, MoSe2, WSe2) flakes. The system can successfully identify highly transparent monolayer hBN, a challenging proposition for other characterization tools. The optical microscope integrated ellipsometer can also map minute thickness variations over a micron-scale flake, revealing its lateral inhomogeneity. The prospect of adding standard optical elements to augment generic optical imaging and spectroscopy setups with accurate in situ ellipsometric mapping capability presents potential opportunities for investigation of exfoliated 2D materials.

Nanopore sequencing identifies a higher frequency and expanded spectrum of mitochondrial DNA deletion mutations in human aging.

Mitochondrial DNA (mtDNA) deletion mutations cause many human diseases and are linked to age-induced mitochondrial dysfunction. Mapping the mutation spectrum and quantifying mtDNA deletion mutation frequency is challenging with next-generation sequencing methods. We hypothesized that long-read sequencing of human mtDNA across the lifespan would detect a broader spectrum of mtDNA rearrangements and provide a more accurate measurement of their frequency. We employed nanopore Cas9-targeted sequencing (nCATS) to map and quantitate mtDNA deletion mutations and develop analyses that are fit-for-purpose. We analyzed total DNA from vastus lateralis muscle in 15 males ranging from 20 to 81 years of age and substantia nigra from three 20-year-old and three 79-year-old men. We found that mtDNA deletion mutations detected by nCATS increased exponentially with age and mapped to a wider region of the mitochondrial genome than previously reported. Using simulated data, we observed that large deletions are often reported as chimeric alignments. To address this, we developed two algorithms for deletion identification which yield consistent deletion mapping and identify both previously reported and novel mtDNA deletion breakpoints. The identified mtDNA deletion frequency measured by nCATS correlates strongly with chronological age and predicts the deletion frequency as measured by digital PCR approaches. In substantia nigra, we observed a similar frequency of age-related mtDNA deletions to those observed in muscle samples, but noted a distinct spectrum of deletion breakpoints. NCATS-mtDNA sequencing allows the identification of mtDNA deletions on a single-molecule level, characterizing the strong relationship between mtDNA deletion frequency and chronological aging.

A multimodal loop closure fusion for autonomous vehicles SLAM

Place recognition and loop closure detection are critical steps in the process of Simultaneous Localization and Mapping (SLAM). Indeed, the ability to determine whether an Autonomous Ground Vehicle (AGV) has returned to a previously visited place is highly important in the context of building a reliable SLAM system. In order to build a consistent global map and to localize the AGV with high confidence in an unknown environment, it is crucial to reduce the cumulative error generated by pose estimation. Although multiple approaches using various data sources have been proposed in order to provide an accurate pose estimation, fewer studies have focused on the integration of a multimodal process to detect loop closure. In this work, we present a novel approach to leverage multiple modalities for a robust and reliable loop closure detection. Our method is based on Similarity-Guided Particle Filtering (SGPF) for the search and validation of Loop Closure Candidates (LCCs). We validate the proposed Multimodal Loop Closure (MMLC) by using two perception modalities based on Bag-of-Words and Scan Context techniques for camera-based and LiDAR-based place recognition, respectively. The efficiency of our method has been evaluated on both KITTI and a self-collected dataset. Compared to the classical loop closure used in ORB-SLAM2, the suggested approach reduces the Absolute Trajectory Error (ATE) by up to 54% and the cumulative error during run-time by up to 62.63%. Finally, 100% of the loops are accurately detected and the ground truth distance between the current pose and the LC is less than 3 m in 98% of the cases.

Does MOED Rhyme with FRUIT? An event-related potential study of cross-language rhyming.

The study aimed to characterize the event-related potentials signature elicited by visual rhyme judgements across two alphabetic orthographies that differ in depth (shallow: Dutch; deep: English) and by spelling-sound consistency in a deep L2-English orthography. Twenty-four Dutch-English bilinguals who varied on measures of L2-English proficiency, made rhyme judgments of semantically unrelated Dutch-English word pairs presented sequentially in the visual modality, while behavioral and electrophysiological responses were recorded. The spelling-sound consistency of target words was varied systematically. Nonrhyming targets elicited a larger N450 amplitude than rhyming targets, indicating sensitivity to mismatching phonology across languages. English target words with consistent spelling-sound mappings elicited less negative N250 amplitudes when preceded by rhyming Dutch primes. Overall, event-related potentials revealed robust responses to phonological mismatch, but subtle responses to spelling-sound inconsistency in L2-English. Results suggest that bilingual readers of a shallow L1 orthography who are immersed in an L1-speaking environment may not tune into the degree of spelling-sound consistency of a deep L2 orthography.

A geometry-aware deep network for depth estimation in monocular endoscopy

Monocular depth estimation is critical for endoscopists to perform spatial perception and 3D navigation of surgical sites. However, most of the existing methods ignore the important geometric structural consistency, which inevitably leads to performance degradation and distortion of 3D reconstruction. To address this issue, we introduce a gradient loss to penalize edge fluctuations ambiguous around stepped edge structures and a normal loss to explicitly express the sensitivity to frequently small structures, and propose a geometric consistency loss to spreads the spatial information across the sample grids to constrain the global geometric anatomy structures. In addition, we develop a synthetic RGB-Depth dataset that captures the anatomical structures under reflections and illumination variations. The proposed method is extensively validated across different datasets and clinical images and achieves mean RMSE values of 0.066 (stomach), 0.029 (small intestine), and 0.139 (colon) on the EndoSLAM dataset. The generalizability of the proposed method achieves mean RMSE values of 12.604 (T1-L1), 9.930 (T2-L2), and 13.893 (T3-L3) on the ColonDepth dataset. The experimental results show that our method exceeds previous state-of-the-art competitors and generates more consistent depth maps and reasonable anatomical structures. The quality of intraoperative 3D structure perception from endoscopic videos of the proposed method meets the accuracy requirements of video-CT registration algorithms for endoscopic navigation. The dataset and the source code will be available at https://github.com/YYM-SIA/LINGMI-MR.

Feature-Based Occupancy Map-Merging for Collaborative SLAM.

One of the most frequently used approaches to represent collaborative mapping are probabilistic occupancy grid maps. These maps can be exchanged and integrated among robots to reduce the overall exploration time, which is the main advantage of the collaborative systems. Such map fusion requires solving the unknown initial correspondence problem. This article presents an effective feature-based map fusion approach that includes processing the spatial occupancy probabilities and detecting features based on locally adaptive nonlinear diffusion filtering. We also present a procedure to verify and accept the correct transformation to avoid ambiguous map merging. Further, a global grid fusion strategy based on the Bayesian inference, which is independent of the order of merging, is also provided. It is shown that the presented method is suitable for identifying geometrically consistent features across various mapping conditions, such as low overlapping and different grid resolutions. We also present the results based on hierarchical map fusion to merge six individual maps at once in order to constrict a consistent global map for SLAM.

UPLP-SLAM: Unified point-line-plane feature fusion for RGB-D visual SLAM

Most of the existing RGB-D simultaneous localization and mapping (SLAM) systems are based on point features or point-line features or point-plane features. However, the existing multi- feature fusion SLAM methods based on the filter framework are not accurate and robust. And the fusion methods based on the optimization framework process different kinds of features separately and integrate them loosely. In the optimization-based framework, how to tightly fuse various kinds of features for achieving more accurate and robust pose estimation has been rarely considered. In this paper, we propose a unified point-line-plane fusion RGB-D visual SLAM method for navigation of mobile robots in structured environments, making full use of the information of three kinds of geometric features. Specifically, it extracts point, line, and plane features using images captured by the RGB-D camera and expresses them in a uniform way. Then, a mutual association scheme is designed for data association of point, line and plane features, which not only considers the correspondence of homogeneous features, i.e., point-point, line-line and plane-plane pairs, but also includes the association of heterogeneous features, i.e., point- line, point-plane and line-plane pairs. Afterwards, the matching errors between homogeneous features and the association errors between heterogeneous features are uniformly represented and jointly optimized to estimate the camera pose and feature parameters for accurate and consistent localization and map building. It is worth pointing out that the proposed unified framework contains two levels. From the system framework perspective, all the main components of the SLAM system, i.e., feature representation, feature association and error function are handled in a unified manner, which increases the accuracy and compactness of the multi-feature SLAM system. From the feature processing perspective, both homogeneous features and heterogeneous features are uniformly used, which provides more spatial constraints on pose estimation. Finally, the accuracy and robustness of the proposed method are verified by experiment comparisons with state-of-the-art visual SLAM systems on public datasets and in real-world environments

Large-Scale LiDAR Consistent Mapping Using Hierarchical LiDAR Bundle Adjustment

Reconstructing an accurate and consistent large-scale LiDAR point cloud map is crucial for robotics applications. The existing solution, pose graph optimization, though it is time-efficient, does not directly optimize the mapping consistency. LiDAR bundle adjustment (BA) has been recently proposed to resolve this issue; however, it is too time-consuming on large-scale maps. To mitigate this problem, this paper presents a globally consistent and efficient mapping method suitable for large-scale maps. Our proposed work consists of a bottom-up hierarchical BA and a top-down pose graph optimization, which combines the advantages of both methods. With the hierarchical design, we solve multiple BA problems with a much smaller Hessian matrix size than the original BA; with the pose graph optimization, we smoothly and efficiently update the LiDAR poses. The effectiveness and robustness of our proposed approach have been validated on multiple spatially and timely large-scale public spinning LiDAR datasets, i.e., KITTI, MulRan and Newer College, and self-collected solid-state LiDAR datasets under structured and unstructured scenes. With proper setups, we demonstrate our work could generate a globally consistent map with around 12 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\%$</tex-math></inline-formula> of the sequence time.