A Deep Learning and Intelligent Recognition Method of Image Data for Rock Mineral and its ImplementationEI北大核心CSCD

Mineral identification is an important task of geological survey, which can be a big challenge in complex geological conditions. In this research, the color and brightness are distinguished using mineral knowledge firstly, which is used to bring out the texture features more obviously. Next, the deep learning model and transfer learning method are applied to build a mineral identification model with enhanced features. Meanwhile, a color model for mineral identification using K-means is also established. Finally, a comprehensive model combining deep learning model and color model is built and a framework of identifying minerals is proposed. In the model establishing process, 6203 images with 19 labels are trained and the top-1 and top-3 accuracy reach 72.2% and 91.2%, respectively. Furthermore, a mineral identification application is developed on the Android system. The application is made up of three modules including mineral knowledge, mineral identification and data storage, which provide real-time recognition, mineral knowledge and data storage. © 2020, Science Press. All right reserved.     

“Three-Component” Digital Prospecting Method：A New Approach for Mineral Resources Quantitative Prediction and Assessment

“Three-component” method consists of three clase-connected aspects: geological anomaly,diversity of mineralization and mineral deposit spectrum. All these three concepts are not new separately, but it is a new approach to combine these three aspects in one single concept for quantitative mineral resources prediction and assessment and it is also the first time to conduct a more detailed study in each aspect. Investigation and clarification of geological anomalies, diversity of mineralization and spectrum of mineral deposits are realized by digitization and quantification of ore forming controlling factors, oreexisting symbols or marks, characteristics of mineralization and regulation of ore-genesis and laws of distribution. These procedures lead to construction of a “digital model“ for mineral resources prediction andassessment.     

Geological and Geochemical Characteristics and Genesis of “Dolomite” at Bayan Obo, China

白云鄂博“白云岩”位于华北板块的北缘,宽沟背斜的南翼.蕴藏着世界最大的稀土矿床,还是大型—超大型铌、铁和钍等矿床赋存母岩.“白云岩”不是层状岩石,无明显的层理和固定的层位,而是一套大小不等,串珠状的”白云岩”带.“白云岩”与围岩呈明显的侵入关系.表现在“白云岩”切割H4石英砂岩、H5板岩和花岗岩脉,并有许多分枝脉插入到H4石英砂岩和H5板岩中.“白云岩”中存在H4石英砂岩的残留顶盖相,捕获了H4石英砂岩和H5板岩的捕掳体,并引起围岩的强烈蚀变.板岩的黑云化,石英砂岩的钠闪石化,花岗岩的碱交代.“白云岩”含有大量铌、稀土、钍等岩浆岩中常见的矿物晶体.硫、碳、氧、锶和铁等同位素组成都具有深源特征.说明白云鄂博“白云岩”不是沉积岩,而是岩浆碳酸岩.     

“Deep-time Digital Basin” Based on Big Data and Artificial Intelligence

正1 Introduction Information technology has been playing an ever-increasing role in geoscience.Sphisicated database platforms are essential for geological data storage,analysis and exchange of Big Data(Feblowitz,2013;Zhang et al.,2016;Teng et al.,2016;Tian and Li,2018).The United States has built an information-sharing platform for state-owned scientific data as a national strategy.National core labs and related database have been established for     

Geochemical properties of Neoproterozoic “cap dolomites” in the Patom paleobasin and problem of their genesis

The characteristic feature of many Upper Neoproterozoic glacial sequences is their “cap carbonates” (CC) resting without visible unconformity upon glaciogenic diamictites. Such an unusual association, peculiar structures and textures, and negative δ¹³C values (approximately −4 ± 2‰) that are atypical of marine carbonates provoked long debates about the nature of these carbonates, which play an important role in the Snowball Earth hypothesis. According to this hypothesis, the Earth was entirely covered by ice during large-scale glaciations, and CC accumulation was related to the global change in geochemical processes. In this work, we discuss data on the chemical and isotopic (C, O, Sr) compositions of CCs, which overlie glacial sediments of the Nichatka and Bol’shoi Patom formations accumulated in different parts of the Neoproterozoic Patom paleobasin (Central Siberia). High concentrations of Fe (up to 6400 ppm), Mn (2320 ppm), and radiogenic Sr (⁸⁷Sr/⁸⁶Sr₀ up to 0.7172) established in CCs indicates a strong influence of the continental flow. Extraordinary Snowball Earth conditions are not necessary for the accumulation of these rocks, geochemical and sedimentological properties of which may be explained by the discharge of thawing waters into partly or completely isolated near-glacier basin, their intermittent freezing, and/or washout of “frozen” carbonates from the surface of thawing glaciers. The peculiar thin-laminated texture of CC may be related to seasonal processes of climatic cycles. They were accumulated in the course of general (relatively long-term) depletion of the atmosphere and hydrosphere in ¹³C, which has nothing to do with the CC formation as a specific type of carbonate sediments. Amplitude and duration of the negative δ¹³C excursion in carbonates associated with the Lower Vendian glacial sediments (665–635 Ma) are appreciably lower than the negative anomaly in rocks of the Zhuya Group that likely correspond to the Shuram-Vonoka Event (∼560−580 Ma ago), which probably marks the crucial point in the Precambrian deglaciation: mass destabilization of methane hydrates and degradation of the Early Vendian psychrosphere in oceans.     

Diamonds and Their Mineral Inclusions,and What They Tell Us: A Detailed “Pull-Apart” of a Diamondiferous Eclogite

For the first time, three-dimensional, high-resolution X-ray computed tomography (HRXCT) of an eclogite xenolith from Yakutia has successfully imaged diamonds and their textural relationships with coexisting minerals. Thirty (30) macrodiamonds (≥1 mm), with a total weight of just over 3 carats, for an ore grade of some 27,000 ct/ton, were found in a small (4 × 5 × 6 cm) eclogite, U51/3, from Udachnaya. Based upon 3-D imaging, the diamonds appear to be associated with zones of secondary alteration of clinopyroxene (Cpx) in the xenolith. The presence of diamonds with secondary minerals strongly suggests that the diamonds formed after the eclogite, in conjunction with meta-somatic input(s) of carbon-rich fluids. Metasomatic processes are also indicated by the non-systematic variations in Cpx inclusion chemistry in the several diamonds. The inclusions in the diamonds vary considerably in major- and trace-element chemistry within and between diamonds, and do not correspond to the minerals of the host eclogite, whose compositions are extremely homogeneous. Some Cpx inclusions possess +Eu anomalies, probably inherited from their crustal source rocks. The only consistent feature for the Cpx crystals in the inclusions is that they have higher K₂O than the Cpx grains in the host.

The δ¹³C compositions are relatively constant at ?5% both within and between diamonds, whereas δ¹⁵N values vary from ?2.8% to ?15.8%. Within a diamond, the total N varies considerably from 15 to 285 ppm in one diamond to 103 to 1250 ppm in another. Cathodoluminescent imaging reveals extremely contorted zonations and complex growth histories in the diamonds, indicating large variations in growth environments for each diamond.

This study directly bears on the concept of diamond inclusions as time capsules for investigating the mantle of the Earth. If diamonds and their inclusions can vary so much within this one small xenolith, the significance of their compositions is a serious question that must be addressed in all diamond-inclusion endeavors.     

Heavy metal distributions in the sediments of “Cattle Pond” on the Dongdao Island of South China Sea: Geochemical and statistical approaches for source tracing

The Dongdao Island of the Xisha atolls is located in the center of the South China Sea. The nearby ambient sea, with thousands of reefs, atolls, submerged reefs and banks, is probably one of the most biologically diverse bodies of water on the planet. Due…