Search

Semantic segmentation is a fundamental technology for autonomous driving. It has a high demand for inference speed and accuracy. However, a good trade‐off between accuracy and latency is yet not present in existing semantic segmentation approaches. Due to the limitation of speed, the authors cannot increase the number of network layers without limit and cannot design modules like in the networks without real‐time. It is a challenging problem how to design a model with good performance under limited resources. To alleviate these issues, in this study, the authors propose a refinement co‐supervision network (RCNet), which is real‐time on a high‐resolution image (1024×2048). The authors first construct the context refinement module, which can provide low computation cost way for obtaining the large receptive field and context information. Furthermore, a boundary co‐supervision mechanism is proposed. It strengthens the optimisation of easily neglected boundaries and small targets. Experimental results reveal that the proposed RCNet outperforms seven representative semantic segmentation methods. [ABSTRACT FROM AUTHOR]

The metasedimentary rocks of the South Liaohe Group are widely distributed along the Jiao‐Liao‐Ji Belt (JLJB) in eastern Liaoning Province. It is important to understanding tectonic setting and evolution of the JLJB by determining its sediment sources and properties of the provenances, sedimentary facies, metamorphic ages, and facies. After the detailed field geological surveys and comprehensive laboratory observation, many high‐grade metapelitic rocks are identified in the South Liaohe Group. The main rock type is sillimanite‐garnet‐cordierite‐biotite plagiogneiss. The zircon U–Pb geochronology of three samples shows that these high‐grade meta‐pelitic rocks are mainly derived from the Liaoji Granites and the Archean basement. The protolith formation ages are between 2.00 and 1.93 Ga. Subsequently, these protolithes underwent high‐pressure metamorphism resulted from collisional orogeny, with peak metamorphism at ca. 1.88 Ga. Then, at 1.88–1.84 Ga, these rocks experienced the retrograde metamorphism and exhumed to the surface during crustal uplift. Geochemical data show that the protoliths are alumina‐rich claystones and graywackes with low maturity, implying near‐source sedimentation. The chemical index of alteration (CIA) values of these samples range from 60 to 80, indicating that the protoliths underwent moderate weathering. The sediment source was mainly felsic components of the upper crust, possibly with the addition of ancient sediments. Various tectonic environment discrimination diagrams combined with previous results on metamorphism and isotope chronology of these metasedimentary rocks in the JLJB indicate that the rocks in the study area originated from an active continental margin with an evolving island arc. [ABSTRACT FROM AUTHOR]

In this letter, we propose a simple yet effective texture descriptor, symmetric mean and directional contour pattern (SMDCP), for texture classification. In particular, first the robust symmetric mean pattern (RSMP) that extracts the sign and amplitude information of the local difference through the neighbourhood average in a new scheme of encoding to further enhance the robustness to noise is constructed. Then a local directional and contour pattern (LDCP) to represent the contour information and direction information of adjacent sampling points is extracted. By concatenating the RSMP and LDCP, a robust and effective texture descriptor (SMDCP) for classification is built. Experimental results reveal that the proposed method obtains significant performance and high discrimination in comparison with 10 representative approaches. [ABSTRACT FROM AUTHOR]

The Palaeoproterozoic tectonic evolution of the Jiao–Liao–Ji Belt (JLJB) in the North China Craton remains controversial, particularly in the absence of adequate evidence for oceanic plate subduction. Here, we report SHRIMP U–Pb zircon ages, whole‐rock geochemistry, and Sm–Nd isotope data on felsic tuffs from the Anshan–Helan–Lianshanguang area in the Liaodong Peninsula which provides important insights into the early tectonic evolution of the JLJB. The zircon SHRIMP U–Pb ages suggest formation of the tuffs crystallized at ca. 2.17 Ga, which coincide with the crystallization of the voluminous Liaoji granitoids in this belt. The felsic tuffs show calc‐alkaline features and are characterized by a wide range of SiO2 (61.89–72.55 wt.%), MgO (1.35–5.80 wt.%), Cr (36.0–67.8 ppm) and Ni (5.88–28.8 ppm) contents, and low TiO2 (0.42–0.64 wt.%) and total Fe2O3 (2.14–5.64 wt.%) concentrations. They also display enrichment in LREE and depletion in Nb, Ta, and Ti similar to the features of continental arc andesites. Their wide variation in whole‐rock εNd(t) (−2.6 to +3.4) combined with the synchronous magmatic activity in this region suggests that they were derived from mixing of mafic (e.g., those in the Liaohe suite) and felsic (e.g., the Liaoji granitoids) magmas, followed by fractional crystallization. In view of the widespread distribution of magmatism, including the Helan felsic tuffs and Liaoji granitoids, and the synchronous sedimentary rocks in the Liaodong Peninsula, we propose that the evolution of the JLJB involved an initial back‐arc extension stage, followed by arc‐continent collision. [ABSTRACT FROM AUTHOR]

Local binary pattern (LBP) is a frequently‐used texture descriptor. Lots of LBP‐variants have been proposed to improve its performance of representing textures. However, most of them ignore the global and neighbour‐difference information of an image texture. In this study, the authors propose a structural difference histogram representation by fusing the segmented structure pattern (SSP), the refined LBP (RLBP) and the neighbour‐difference pattern (NDP) for texture classification. Particularly, the segmented structure, which contains the global contour information of an image texture, is first constructed to compute its SSP histogram feature. Simultaneously, the RLBP is defined to represent the local texture information. Furthermore, the NDP is presented to describe differences between neighbours of a centre pixel in the local patch of texture images. Experimental results on Brodatz and Columbia‐Utrecht reflectance databases indicate that the proposed method can achieve the satisfactory classification accuracy compared with several representative methods. [ABSTRACT FROM AUTHOR]