Search Documents

By Paul Zinner

An exhaustive program of geologic studies and mapping has been conducted by the Geological Survey in the Iron River district of Michigan since May 1943. Means of contributing to the success of these studies were sought by the Bureau of Mines, and eventually a program involving geophysical studies and core-drilling was undertaken in the area. All of the work was done either to develop data that would contribute directly to geologic studies or to investigate geophysical methods that might be used as a primary means for the discovery of minable iron ore.

Jan 1, 1949

By W. L. Dalke

The Minntac complex is a facility for production of iron ore pellets located on the Mesabi Range at mountain iron, Minnesota. It consists basically of a mine, crushing plant, concentrator, and a pelletizer producing twelve and one-half million tons of pellets annually. This paper deals with some of the aspects of the grinding circuits; but prior to that discussion, it might be well to briefly describe the flow of material in the operation. The crude ore, containing from 20% to 25% magnetic iron, is mined by 14-yard electric shovels. .the blasted ore is loaded directly into 50-yard side dump railroad cars and hauled to the crushing plant. Crushing is done in three stages, originally, two 60" primary crushers; four 7-foot standard secondary crushers; and six 7-foot short head tertiary crushers were installed, when the plant was expanded to double its capacity, five secondary and ten tertiary crushers were added. These were duplicates of the original 7-foot standard and short head crushers used in the original construction, there was no expansion of the primary crushers but a one-half-million-ton surge pile of tertiary crusher feed was added.

Jan 1, 1976

By C. Hu

A MOS capacitor hydrogen sensor array is reported for hydrogen gas distribution mapping. Si(100) was used as substrate for film deposition. After silicon oxide growing on the silicon surface with dry thermal oxidation, palladium film was sputtered on silicon oxide as hydrogen-sensitive gate material in MOS structure. The 3x3 sensor array was patterned on silicon substrate and packaged in one chip. Sensors' response was test with impedance analyzer. The sensors in array were calibrated in the hydrogen concentration range from 10ppm to 10000ppm. The 2D hydrogen concentration distribution was obtained with the calibrated sensor array.

Jan 1, 2005

By M. J. Ramos

Without an exclusive definition of administration in business, we could propose to define it as performance by management to adopt or implement established business formulas. Administration could also be defined as a universal process that organizes people, and achieves, through efficiency, common objectives and goals. Humanity developed some form of this process from the beginning of civilization. To imply that something is new in administration might be considered specious or naïve. Moreover, what is new in management? Administration of business is not new to the world. To prove this statement, we need only ask this question: Does the architecture and monumental workmanship of the past civilizations of ancient Egypt, Rome, and the Incas demonstrate management without administration and without definition of objectives? The answer must be a resounding ?NO!? In spite of this acknowledgment and for the purposes of this presentation, I will use the word ?new? in business and hope that the reader will follow my reasoning. The mining industry has unique characteristics in respect to other industries. It is multi-disciplinary, has a single purpose (majority for one metal), has two players (man and nature), and it involves high risks for generating environmental imbalance. Mining exploits non-renewable and finite resources, creates an unstable job market (work hard to have no work) and a speculative economic market with unpredictable prices ? a one-time payment for a finite resource with multiple recycling. On top of all that, the industry still operates within obsolete petroleum-based technology, utilizing practices that were designed two centuries ago. Today, we confront the challenges of the millennium with outmoded practices and management. Ordinarily, several styles of administration are standardized into a FRAGMENTED vertical system of organization. Mining has rigid objectives in production, with profitability as the focus from the beginning to the end. Citing the concept of safety, as an important facet of production is bombastic since, in truth, mining is an experimental industry that consistently reports high numbers of fatal accidents. In mining, we designed a process to work according to our objectives, demanding attention and supervision of all levels of management with a halo of parameters that are not assimilated into the design. We continue to perform a rigid style of management in an industry that is still not responsive to variations of the market ? losses in particular. The existing design process generates loss without capitalization and without response to said loss. This perception is reflected in a mindset that is only too ready to ignore mining liabilities. The reality of the enterprise has created an immense pressure from top to bottom in the hierarchy to do more with less. It is for all of these reasons that I propose to set forth a new essence for administration that encompasses an alternative vision and style in mining: Management by Loss. MANAGEMENT BY LOSS Management by Loss (MBL) is a systematic, integral process organized to quality control operation and the human profitability within the industry. It permits focus on achievable goals by utilizing the best practices (Mining Open-closing). Identification and characterization of loss in the process determine how objectives are defined, controlling the time limits to execute based on constant balance of mass, energy, water, and ?wastes? in each stage of the industry, in order to reach final objectives. The concept of loss can be defined by various criteria. Generally, it is the mass that could not make or arrive at its destination from a designed process or simply, it could not maintain balance. Otherwise, the design of the process describes the final point of the resource. However, those resources could not make their destination because of failure to use the best design, resources and TIME. The loss has typical characteristics such as Identifiable, classifiable, measurable, manageable, controllable, valuables, negotiable, avoidable, reportable, and monitoring. In the traditional scheme, these losses are not registered; they are made exclusions. We are trapped within a system, which was designed to obtain objectives of the industry in a specific timeframe. In the present style of administration, the focus is total - our emphasis in this treatise is in the loss. The current traditional report form contains only the most positive achievement of objectives, only the best of the exercise, humanizing numbers, without touching what we left behind or what we were waiting for to assimilate as a ?surprise account?(new account of liabilities). Management by loss is a method that redefines the concept of waste (waste zero) for controlling the mine operation under the equilibrium of the ecosystems. This method confronts the problems that impact our industry in this millennium. Understanding and quantifying losses (reserves, resources, time, energy, water, soil, materials, equipment, tasks, labor, ecosystems, etc) is a priority mandate for the industry and demands the application of a ?new ? style of administration by seeking to maintain balance instead of the disequilibrium that mining now creates in nature. It is necessary to answer questions such as: How can we minimize loss? How can the loss be assimilated during the life of the mine? What is the vision of mining in respect to the generation of waste? How can we convert liabilities into assets? In addition, the control of loss is determined in the achievement of the plan. Loss occurs because of a failure of the process in the original plan. Still we do not plan to fail but we fail to plan appropriately. It is a remarkable thing to accomplish and compel those objectives described in the original plan when that plan is designed with a determined budget. If not, we lose time and, of course, capital. Moreover, in Management by Loss (MBL), we can find answers to the following questions: What is the multiplier of productivity in mining? What are the influences of identification and characterization of loss and what is the level of efficiency of energy in order to achieve waste zero (effective cut off grade)? The challenges of mining in this millennium validate this new style of administration. Finally, it is important to point out that mining is entering into an era that has different rules, but the industry is still run by companies functioning with objectives from two centuries ago. They are trying to confront issues with concepts that are not designed for new tasks. Therefore, I ask you to consider this new style of administration of business under a vision of how the industry should be organized and managed, with an attitude of balance toward nature and social responsibility.

Jan 1, 2009

By Ira Brotman

The Virginia Port Authority and the U.S. Army Corps of Engineers are partnering to construct the Craney Island Eastward Expansion (CIEE) to meet the dual purpose of extending the life of Craney Island as a dredged material management area and providing land for a state-of-the-art marine terminal. The over 500-acre (200-hectare) expansion provides additional upland dredged material capacity for the Hampton Roads Harbor by constructing perimeter dikes, that once filled, will provide land for the development of a container terminal. In the winter of 2010/2011 the first stage of sand fill was placed, followed by the installation of approximately 12 million linear feet (MLF) (3.6 million linear meters, MLM) of marine PVDs. With over 36 MLF (10.9 MLM) of PVD required to complete dike construction the project will be one of the largest PVD installation project in the U.S. his paper details the challenges associated with the construction of the cross dikes, in particular the design, testing, installation and monitoring of prefabricated vertical drains (PVDs) installed overwater to depths of over 130 feet (40 meters) for foundation improvement in the insitu very soft Norfolk Clay. With expected settlements of 20?30 ft (7-10 m) and 30 percent average strains, the discharge capacity of the PVDs was of particular concern. Laboratory data is presented showing the magnitudes and variability of discharge capacity under different confining pressures in both a straight and buckled condition. The method of installation and monitoring of the marine PVDs in open water included developing a specialized barge and sophisticated positioning and data collection system to verify installation location, depth of penetration and track quantities. An observational method, including material testing, detailed surveys, periodic in-situ testing, and instrumentation, was utilized by the project team to assess the project?s progress and allow for validation of consolidation and strength gain rates and magnitudes.

During the life of a mining project in Papua New Guinea (PNG), the greatest impact is on the people of the project area and the most intense, day-to-day contact is between these people and the project's field staff. Yet formal structures at present emphasise contact between much higher levels of national government and company personnel. Whilst understandable, this is unfortunate particularly since, given the nature of PNG politics, villagers do have a powerful influence on project success, as recent dramatic events on Bougainville have indicated. A greater role for local views and greater attention to such views by other parties is clearly required. But to achieve a successful marriage between informal village politics and the formal structures of investment procedures will require considerable patience and adjustment. This should be reflected in the cost of project development and be regarded as being as of a significance equal to that currently accorded to the technical aspects of project development.

Jan 1, 1990

By William G. Fischer

This paper attempts to close the gap between extensive rock testing with field measurements and the mine operator's need for safe mine layouts and alternatives. It was found that many of the variables that influence mine geometry and worker safety are relatively unimportant while others are critical. The development of empirical relationships which tie the critical variables to the "safe useful life" of a trona mine layout and observations made in 140 production areas where the variables were researched over a fifteen year period are discussed. The concepts presented may have applicability in mine design for other bedded deposits.

Jan 1, 1984

By K. R. Bixby

OPENING of numerous stripping operations in Ohio, Pennsylvania, and other districts, particularly outside the Middle West and Southwest where the large-scale stripping mines predominate, holds the limelight for the year 1940. Several factors are responsible for this expansion. One is the scarcity of untested and undeveloped properties in this central field. Another has been the development of earth-moving and preparation equipment suitable for these smaller tonnage tracts. Draglines of the Walker type, small long-range .hovels, and large-capacity tractor- drawn scrapers have to be considered. Wash boxes, designed for light tonnages, costing a few thousand dollars, have made their contribution.

Jan 1, 1941

By Tanya J. Gallegos, Mojisola A. KunleDare, James V. Jones III, Charlotte E. Riggs

The United States Geological Survey (USGS) provides scientific information for the Department of Interior and the nation, consistent with its original mission expressed in the Organic Act of 1879 (43 U.S.C. 31): “the classification of the public lands and examination of the geological structure, mineral resources, and products within and outside the national domain.” Legislation such as the Energy Act of 2020 and the 2022 Infrastructure Investment and Jobs Act (43 USC 31l) and recent executive actions (Executive Orders 14154 , 14153, 14241 Secretary’s Orders 3417, 3418, 3422, 3436) underscore the importance of mineral resources and focus USGS activities on mapping and assessing mineral resources, with a particular focus on those presently identified as critical, both in ground and above ground in mine wastes. This article reviews selected activities and accomplishments by the USGS Mineral Resources Program related to critical minerals in 2025. Highlights include a new List of Critical Minerals, a first-ever national mine waste inventory, international minerals partnerships, mineral resource assessment advancements, and national data collection activities and outcomes of the Earth Mapping Resources Initiative (Earth MRI). The selected contributions are not comprehensive but are intended to demonstrate USGS leadership in critical mineral mapping and assessment, the importance of domestic and global partnerships, and the breadth of research activities that are responsive to national needs and priorities.

May 1, 2026

By T. F. Dumm, O. Trass, V. R. Koka, L. G. Austin

"The size distributions of carefully screened 2 size fractions of ground materials were determined using a Microtrac laser dif-fractometer. The size distributions were fitted to the empirical function P(x) = 1/[1 + (x50/x)1] where P(x) is cumulative mass fraction less than Microtrac size, x. The standard deviation of A for a 50 second test time was approximately 0.5, so that seventeen test times give a mean with an estimated error within t 0.25 (95% confidence level). Values of were 5.44, 4.82, 4.40 for a coal ground under different conditions, indicating statistically different shape distributions; the average value for mica was 2.91 owing to the high aspect ratio of the particles. The ratio of x50 to the geometric mean sieve size was 1.2 for the coal and 0.70 for mica.1 IntroductionKoka and Trass have presented evidence [1,2] that materials ground in the Szego mill [3] give different shapes depending on the milling conditions. The variations in shape were deduced qualitatively by viewing SEM photomicrographs of selected 1/2 screen fractions. This paper quantifies their observed differences in shape by using measurements of the size distributions of screen fractions examined in a laser diffractometer [4] (Microtrac, Model 7991). The basic concept is that particles of 'a given ""screen"" size will give a wider distribution of ""Microtrac"" size when the particles are flaky with high aspect ratios, and a narrower distribution when the particles approach the cubical and spherical geometries.2 TheoryAustin and Shah [5] have shown that coal screened within a 1. 1/2 upper-to-lower sieve size range will give rise to a much wider distribution of Microtrac size. They found that the screen interval of irregular fragments of comminuted coals or quartz gave Microtrac size distributions which could be described by the simple function"

Jan 1, 1988

By Sunil Bodakunta, Sreenivasa Rao Islavath

The abstract highlights the importance of determining the optimum line of extraction in continuous miner workings for improved safety and productivity. The study investigates three extraction techniques (diagonal, straight line along strike, and straight line along dip-rise) using 3D numerical modeling. The dip-rise extraction technique is found to be optimal due to lower stress and displacement, leading to a higher safety factor. The research also develops a continuous miner stability index (CMSI) to predict working face stability based on mining depth and pillar width-to-height ratio.

Apr 28, 2022

By Mark A. Ackerson, Kenneth E. Hjelmstad

An ideal fire warning alarm system for underground mines would be low cost, convenient, fast, reliable, and able to warn all underground workers. Present warning systems, such as phones, messengers, and stench, fail one or more of these criteria. The U.S. Bureau of Mines may have devised the ideal system. The one way communication system employs a large loop antenna and transmitter to create a 630 hertz (Hz) electromagnetic field to send information through-the-earth to cap lamp battery mounted micro- receivers worn by underground miners. Field tests of prototype equipment in 1986 resulted in through-the- rock signal transmission of over 762 m (2500 ft). Subsequent hardware upgrades for more recent tests resulted in transmissions of over one mile. Recently developed receiver circuitry employs phase- locked-loop detection circuitry responsive only to a specific series of pulses of given time duration, that eliminates false alarms from electromagnetic noise emitted by mine equipment, power lines, or lightning. A multichannel system can relay standard messages to miners

By P B. Swart, P D. Erskine

Despite an increasing focus by media, communities and activists driving an elevated regulatory focus in Australia and indeed some other mining jurisdictions, the mining industry globally has had limited successful mine lease relinquishments and transfers to support a next land use.Critique from anti-mining lobbyists frequently uses the large estimates from historically abandoned mine sites to influence risk-averse regulatory reforms to improve environmental performance. The mine rehabilitation regulatory processes in Australia, and specifically in the State of Queensland, are thus largely in place to prevent mining operators from abandoning their obligations.Regulatory responses to the lack of mine rehabilitation and closure performance in the last five to ten years have primarily focussed on compliance and enforcement of rehabilitation conditions in environmental approvals. Additionally, financial assurance measures have been tightened to protect against abandonment of rehabilitation obligations by the Industry, including the Chain of Responsibility legislation (Queensland Government, 2016b) in April 2016, to hold to account those who have benefitted from mining and processing/refining projects, and then avoid their environmental obligations.CITATION:Swart, P B and Erskine, P D, 2018. Are our closure planning and regulatory processes adequate to identify, quantify and appropriately transfer residual risks after successful mine rehabilitation?, in Proceedings Life-of-Mine 2018, pp 155–158 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 25, 2018

By R D. Lama, A Misra

One of the major factors restricting an increase in the productivity of a longwall mine is lower productivity obtained in a development panel. There has been a significant improvement in gate road development, particularly in mines requiring high density roof and rib bolting within 2 m of the face. Advance rates have increased from an average of 3 m to 5 m per shift in 1985 up to an average of 5 m to 10 m per shift in 1991. In mines at shallow depths with reasonably good conditions the increase has not been that significant. Advance rates have increased from an average of 7 m to 12 ni per shift in 1985 to an average of 12 m to 15 m per shift in 1991. Compared to the United States' (US) development rates, these are much lower. Extensive research work has been carried out to minimise cutting, wheeling, support setting time, and to optimise pillar length to achieve maximum advance rates. Pillar optimisation is very important in order to expedite adverse rates. This paper investigates various development rate technologies, both present and future, and optimises pillars to improve these dev elupment rates. Pillar optimisation is based upon time and motion studies conducted at mines in the Illawara region of New' South Wales, Australia, and the basic tools of arithmetic.

Jan 9, 1996

By D. Christofferson

This final report documents the cooperative efforts of the Engineered Systems Division and the Mining Equipment Division of FMC Corporation during performance of Bureau of Mines Contract H0252091. Improved operator compartments incorporating overhead canopies, logical arrangements of controls, and design features offering greater operator comfort were developed and retrofitted to the FMC 320A roof bolter and 6L shuttle care The vehicles were evaluated in underground operations at separate mine sites. The successful evaluation has resulted in incorporation of the designs as standard features of the 320A roof bolter, and the 5L, 6L, and 10L shuttle cars.

Jan 1, 1978

By Y Cai, B Hebblewhite

In this paper, the Australian thick coal seam underground mining potential is highlighted and the characteristics are investigated by analysing a thick seam database. It is clear that thick coal seam resources should be mined with a high recovery, productive, efficient as well as a safe and reliable mining method. Longwall Top-Coal Caving (LTCC) is the most recent thick coal seam mining method which has been developed and extensively used in China for the last ten years. The successful applications indicate that this method has a very high recovery and productivity rate and less development cost in comparison with any other thick coal seam mining method. One of the aims of the current study is to seek potential for the application of LTCC as a thick seam mining method in Australia. In this paper, preliminary results of investigation on Australian thickseam potential are given and results of data analysis for potential LTCC seams in Australia are described. The outcome of this study shows a great promise for the application of LTCC in Australian coal industry.

Jan 1, 2002

By Kumar P. Dharmasena, Haydn N. G. Wadley, Bill W. Choi

"Non-contact eddy current sensing methods have been developed and applied to Cd0.955Zn0.045Te crystal growth to obtain a better understanding of the melt, solid nucleation, and the growth process. The application of this sensor approach relies on a large difference in liquid and solid electrical conductivities near the melting point and requires knowledge of the relationships between the electrical conductivity, melt composition, and temperature. Using a multifrequency encircling eddy current sensor, the electrical conductivities of three different Cd1-xZnxTe (x=0.00, 0.045, and 0.08) alloys were measured as a function of temperature in a laboratory-scale vertical Bridgman furnace. The measured solid and liquid electrical conductivities were then used in electromagnetic finite element models to analyze the multifrequency responses of eddy current sensor designs during the simulated growth of a crystal. These eddy current sensors were fabricated and installed in a vertical Bridgman furnace and used to monitor many Cd0.955Zn0.045 Te growth runs. One sensor was placed in the vicinity of the ampoule tip to detect undercooling/spontaneous nucleation and two other sensors (""absolute"" and ""differential"" designs) placed to monitor the movement/curvature of the liquid-solid interface during crystal growth. The sensor data was used to characterize the initial melt state, detect the onset of nucle-ation, determine the growth velocity, and identify the shape of the interface.IntroductionSingle crystal CdTe and its solid solution Cd1-xZnxTe alloys (0.03<x<0.05) are used as infrared transparent substrates for infrared focal plane array (IRFPA) detectors [1-6] and as solid state y- ray detectors [7-9j. Cd1-xZnxTe substrates are normally ""mined"" from large polycrystalline boules grown by the unseeded directional solidification of Cd1-xZnxTe melts using either a vertical or horizontal Bridgman process [10-14]. The vertical Bridgman method is particularly popular for growing CdTe and its related compounds. In this process, a Cd1-xZnxTe ingot is produced by first melting a charge in the hot zone of the furnace and then vertically translating the furnace (with its associated axial temperature profile) relative to the stationary crucible to cause directional solidification. In spite of many experimental efforts to investigate the relationships between material purity, stoichiometry, the controllable growth parameters (furnace temperature profiles, ampoule geometries/materials, furnace translation rates and starting position) and the resulting material characteristics (grain structure, dislocation densities, IR transmission, macro-segregation etc.), the yield of Cd1-xZnxTe of a quality suitable for large area (e.g. 4 cm x 6 cm) substrates remains low (<10%). Since much of this poor yield is directly associated with the crystal growth process (e.g. melt stoichiometry, solidification velocity, interface shape, temperature gradients, cooling rate, etc.) intensive efforts are under way to improve this technology."

Jan 1, 1997

By Hiroshi Ono

"By newly developed ICP method, trace amounts of As(III), As(V) .,Sb(III) and Sb(V) were precisely determined within accuracy of sub ng/ml. As(III) and Sb(III) were separated from As(V) and Sb(V) by extracting into ·xylene as diethyldithiocarbamate complexes using diethylammonium diethyldithiocarbamate (DDDC), and then back-extracted into nitric acid for a quantitative determiruition by ICP-AES equipped with hydridegeneration system The recoveries of As(III) and Sb(III) were satisfactory throughout the extraction with 0.04mol/l DDDC in xylene from lmol/l hydrochloric acid solution and the back-extraction with 7 mol/l nitric acid. As(V) and Sb(V) were reduced to As(III) and Sb(III) by ascorbic acid and potassium iodide, and their quantities were determined as mentioned above. The detection limits were found to be 0.58ng/ml for As(III), 0.63ng/ml for Sb(III) using 50ml of a sample. The proposed method was applied to analysis of the real sample.Introduction In drain, arsenic(AS) concentration is prescribed below 0.lmg/l, so As must be removed. As is existed at .the form of As(III) or As(V) in solution. For t1Jeir difference of chemic;µ property, it is not easy to remove As in solution. So it is an urgent necessity to develop for determining concentration of As(III) and As(V). Antimony( Sb) has resemblance .chemical property to As. So Sb was experimented the same as As.The separation and determination of As(III) and As(V) have been carried out using several methods, such as ICP after copreclpitating separation, HPLC, direct hydride generation atomic absorption. However these methods have some problems. One of main problems is that HPLC and direct hydride generation atomic absorption methods are influenced by coexisting elements."

Jan 1, 2000

By S Nowosad, A Binder, O Langefeld, M Bothe-Fiekert, F Apollo

The mining industry faces complex and urgent challenges driven by an increasing demand for raw materials, and a socially demanding globalised world. The extraction of primary raw materials is today more important than ever as one of the cornerstones towards a sustainable future since the increasing demand for the raw materials required for the energy transition. In the last three years, leading countries as USA, China, Canada, Australia, and the European Union (EU) have developed and adopted strategies to enhance domestic extraction capabilities, diversifying sources, fostering innovation in recycling, and promoting international collaboration related to critical raw materials. On 2022, the Inflation Reduction Act was signed into law in the USA, in June 2023, the Australian Critical Minerals Strategy 2023–2030 was published and in March 2024, the Council of the EU adopted the European Critical Raw Materials Act. On the other hand, raw materials extraction is facing higher environmental and safety standards and regulations, as well as scrutiny from the broader public. Regrettable environmental accidents as tailing dam failures or acid mine drainage occurrences have led to a general distrust of mining activities leaving mining companies with the dual imperative of increasing output while implementing sustainability, decarbonisation, and circularity as a core value in their production processes. This challenging environment offers the opportunity to enhance ‘blue mining’ (BM) as a sustainable holistic approach that integrates the principles needed to plan sustainable, reliable, and responsible mines for today’s and future generations by implementing the principles of circular economy (CE). This requires a fundamental shift from the traditional linear economy approach to a sustainable, circular economy strategy. At its essence, BM encapsulates four crucial aspects – energy, ergonomics, water, and circularity and integrates these four aspects into integrative mine planning, defining clear goals and objectives from the ‘blueprint’ of a mine towards a beneficial post-mining activity. Its implementation can lead to maximisation of resources beyond the life of the extractive mine and sustainable use of materials by integrating the principles of circular economy and multiuse mine, see Figure 1. These concepts complement and recognise the need to minimise environmental and social impacts while advancing economic development. Finally, this contribution explains the requirements and challenges of moving from linear to dynamic planning approaches under the BM concept. This way, BM stands as an approach that can contribute to the mining industry receiving necessary approvals and support from society leading to a long-term transformation of the industry’s image.

Sep 1, 2024

By E Nettleton, A Melkumyan

Identification of geology and mineralogy in situ on the mine face is essential if mining processes are to be automated for tasks such as exploration, grading and reconciliation. Hyperspectral data acquired from field based platforms are ideally suited to these tasks, and are set to revolutionise the automated mining industry. Changing conditions of solar illumination introduces variation into the hyperspectral signal, which is unrelated to geology, complicating analysis of data. Thus, algorithms must be able to deal with changing illumination conditions and operate in a consistent and probabilistic framework, which is a prerequisite for data fusion with other sensors. To produce accurate classification maps of geology, algorithms must be developed that are effective under different conditions of illumination.The efficacy of three algorithms to classify hyperspectral data for geology mapping on vertical mine faces was assessed. These algorithms are the spectral angle mapper (SAM), and two machine learning methods operating within a fully probabilistic Gaussian process (GP) framework û the squared exponential (SE) and the observation angle dependent (OAD) covariance functions (kernel). Two experiments were done to assess the classification performances. The first determined the performance of the algorithms by classifying hyperspectral data of known rock samples acquired using a field spectrometer. The GP-based methods had a similar performance, with average accuracies and F-scores of 97 and 88 per cent (GP-OAD) and 95 and 81 per cent (GP-SE), respectively. Average accuracies and F-scores for SAM were significantly smaller (80 and 40 per cent respectively). The algorithms were then applied to hyperspectral imagery acquired from a vertical mine face at different times of the day, under very different conditions of illumination. In contrast with experiment&nbsp;1, SAM and GP-OAD gave results which were superior to GP-SE and were entirely consistent with geology mapped in the field. Compared with SAM, GP-OAD was less sensitive to the effects of shadow and produced classified images that were less noisy. The GP-SE method failed to delineate even gross changes in geology on the mine face. These results show conclusively that in order to use independent training data (spectral library) to classify imagery acquired under a different set of conditions, algorithms must take into account variability in incident illumination. Measures based on spectral distance are inferior to those based on spectral angle.The OAD kernel within the GP framework (GP-OAD) is superior to the other two methods because:it provides a probabilistic framework;it is insensitive to variability in illumination within and between imagery acquired under different conditions; andit can identify and map geological zones on a vertical mine face using an independent spectral library, without a priori knowledge.

Sep 26, 2011