CMANet: Cross-Modality Attention Network for Indoor-Scene Semantic Segmentation.

Indoor-scene semantic segmentation is of great significance to indoor navigation, high-precision map creation, route planning, etc. However, incorporating RGB and HHA images for indoor-scene semantic segmentation is a promising yet challenging task, due to the diversity of textures and structures and the disparity of multi-modality in physical significance. In this paper, we propose a Cross-Modality Attention Network (CMANet) that facilitates the extraction of both RGB and HHA features and enhances the cross-modality feature integration. CMANet is constructed under the encoder-decoder architecture. The encoder consists of two parallel branches that successively extract the latent modality features from RGB and HHA images, respectively. Particularly, a novel self-attention mechanism-based Cross-Modality Refine Gate (CMRG) is presented, which bridges the two branches. More importantly, the CMRG achieves cross-modality feature fusion and produces certain refined aggregated features; it serves as the most crucial part of CMANet. The decoder is a multi-stage up-sampled backbone that is composed of different residual blocks at each up-sampling stage. Furthermore, bi-directional multi-step propagation and pyramid supervision are applied to assist the leaning process. To evaluate the effectiveness and efficiency of the proposed method, extensive experiments are conducted on NYUDv2 and SUN RGB-D datasets. Experimental results demonstrate that our method outperforms the existing ones for indoor semantic-segmentation tasks.

Retrofit and application of pulverized coal burners with LNG and oxygen ignition in utility boiler under ultra-low load operation.

An oxygen-rich and low NOx burner integrated with liquefied natural gas (LNG) was proposed to address unstable combustion and high NOx emissions from a 330 MW subcritical boiler under ultra-low load operation in China. To assess the effectiveness of the retrofit, Chemkin and Fluent softwares were utilized to construct a new NOx model and calculate NOx generation, based on the combustion of pulverized coal gas and LNG. Further, an eddy dissipation concept (EDC) model, which can reflect detailed chemical reactions, was applied to calculate gas-phase reactions in the furnace. The results showed that when performing the deep peak shaving after the retrofit, the combustion in the furnace was stable under 50% or more load, and NOx emission level at the furnace outlet was lower than 350 mg/m3 (6% O2 content, dry basis). Under 25% load, the oxygen-rich burner integrated with LNG was applied, and the pulverized coal flow entered the furnace in a state of high-intensity combustion, which effectively promoted the stability of combustion in the furnace. The reductive combustion state with reductive free radicals generated by LNG decomposition inhibited NOx formation. Consequently, NOx emissions from the furnace outlet decreased from 380 mg/m3 to 316 mg/m3.

Automatic Registration of Terrestrial Laser Scanning Point Clouds using Panoramic Reflectance Images.

This paper presents a new approach to the automatic registration of terrestrial laser scanning (TLS) point clouds using panoramic reflectance images. The approach follows a two-step procedure that includes both pair-wise registration and global registration. The pair-wise registration consists of image matching (pixel-to-pixel correspondence) and point cloud registration (point-to-point correspondence), as the correspondence between the image and the point cloud (pixel-to-point) is inherent to the reflectance images. False correspondences are removed by a geometric invariance check. The pixel-to-point correspondence and the computation of the rigid transformation parameters (RTPs) are integrated into an iterative process that allows for the pair-wise registration to be optimised. The global registration of all point clouds is obtained by a bundle adjustment using a circular self-closure constraint. Our approach is tested with both indoor and outdoor scenes acquired by a FARO LS 880 laser scanner with an angular resolution of 0.036° and 0.045°, respectively. The results show that the pair-wise and global registration accuracies are of millimetre and centimetre orders, respectively, and that the process is fully automatic and converges quickly.

Primitive fitting based on the efficient multiBaySAC algorithm.

Although RANSAC is proven to be robust, the original RANSAC algorithm selects hypothesis sets at random, generating numerous iterations and high computational costs because many hypothesis sets are contaminated with outliers. This paper presents a conditional sampling method, multiBaySAC (Bayes SAmple Consensus), that fuses the BaySAC algorithm with candidate model parameters statistical testing for unorganized 3D point clouds to fit multiple primitives. This paper first presents a statistical testing algorithm for a candidate model parameter histogram to detect potential primitives. As the detected initial primitives were optimized using a parallel strategy rather than a sequential one, every data point in the multiBaySAC algorithm was assigned to multiple prior inlier probabilities for initial multiple primitives. Each prior inlier probability determined the probability that a point belongs to the corresponding primitive. We then implemented in parallel a conditional sampling method: BaySAC. With each iteration of the hypothesis testing process, hypothesis sets with the highest inlier probabilities were selected and verified for the existence of multiple primitives, revealing the fitting for multiple primitives. Moreover, the updated version of the initial probability was implemented based on a memorable form of Bayes' Theorem, which describes the relationship between prior and posterior probabilities of a data point by determining whether the hypothesis set to which a data point belongs is correct. The proposed approach was tested using real and synthetic point clouds. The results show that the proposed multiBaySAC algorithm can achieve a high computational efficiency (averaging 34% higher than the efficiency of the sequential RANSAC method) and fitting accuracy (exhibiting good performance in the intersection of two primitives), whereas the sequential RANSAC framework clearly suffers from over- and under-segmentation problems. Future work will aim at further optimizing this strategy through its application to other problems such as multiple point cloud co-registration and multiple image matching.

Global registration of subway tunnel point clouds using an augmented extended Kalman filter and central-axis constraint.

Because tunnels generally have tubular shapes, the distribution of tie points between adjacent scans is usually limited to a narrow region, which makes the problem of registration error accumulation inevitable. In this paper, a global registration method is proposed based on an augmented extended Kalman filter and a central-axis constraint. The point cloud registration is regarded as a stochastic system, and the global registration is considered to be a process that recursively estimates the rigid transformation parameters between each pair of adjacent scans. Therefore, the augmented extended Kalman filter (AEKF) is used to accurately estimate the rigid transformation parameters by eliminating the error accumulation caused by the pair-wise registration. Moreover, because the scanning range of a terrestrial laser scanner can reach hundreds of meters, a single scan can cover a tunnel segment with a length of more than one hundred meters, which means that the central axis extracted from the scan can be employed to control the registration of multiple scans. Therefore, the central axis of the subway tunnel is first determined through the 2D projection of the tunnel point cloud and curve fitting using the RANSAC (RANdom SAmple Consensus) algorithm. Because the extraction of the central axis by quadratic curve fitting may suffer from noise in the tunnel points and from variations in the tunnel, we present a global extraction algorithm that is based on segment-wise quadratic curve fitting. We then derive the central-axis constraint as an additional observation model of AEKF to optimize the registration parameters between each pair of adjacent scans. The proposed approach is tested on terrestrial point clouds that were acquired in a subway tunnel. The results show that the proposed algorithm is capable of improving the accuracy of aligning multiple scans by 48%.