Hyperspectral Remote Sensing Technology in Civil Engineering - Coursework Example

Hyper spectral Remote sensing technology in civil engineering Department admission: Place of admission: Hyper spectral remote sensing technology is basically a developing technology in the field of remote sensing and Geographical Information systems (GIS). Its development started sometimes in the 1980s and over the last two decades, this field has seen great developments in terms of the data collection techniques, data processing and analysis software and the integration in other fields in order to solve environmental and engineering problems. The concept of spectroscopy has a rich history since it dates early in the century when scientists employed this technology in geology and determination of minerals. Today, hyper spectral remote sensing technology is widely used in various fields including agriculture, geology, military personnel and also engineering and construction. This research paper will be explaining and analyzing the engineering applications of hyper spectral imaging technology, which detects street conditions and traffic flow on the highways and the roads. Taking into consideration the importance of these infrastructural conditions to the economic world, this research will be concentrating on how hyper spectral imaging technology can be used in pipelines that transport oil and natural gas. In this research, it is crucial to understand how hyper spectral remote sensing technology can help engineers to discover whether or not there are any curves or rusting within the transporting pipes. In addition, this developed technology allows engineers to also discover and diagnose whether or not there are any issues on the road. Table of contents List of illustrations…………………………………………………………………4 Introduction…………………………………………………………………………5 Discussion……………………………………………………………………………6 Summary ……………………………………………………………………………..15 References……………………………………………………………………………..16 List of illustrations a) Figure 1: An analyzed agricultural area in California……………………………5 b) Figure 2: The components of a hyper spectral remote sensing imager……………6 c) Figure 3: Illustration of CCD output………………………………………………7 d) Figure 4: Illustration of a hyper spectral cube……………………………………..7 e) Figure 5: The concept of hyper spectral imaging…………………………………..8 f) Figure 6: AVIRIS image of Fairview area with distribution of roads……………..9 g) Figure 7: A road condition survey sheet……………………………………………9 h) Figure 8: Comparison of road survey……………………………………………….10 i) Figure 9: Spectral effects of asphalts ageing………………………………………11 j) Figure 10: Results of the ANOVA analysis………………………………………12 k) Figure 11: A pipeline monitoring system…………………………………………12 l) Figure 12: Satellite detected encroachment………………………………………....13 m) Figure 13: Supervised classified image of the leakage area………………………..13 Introduction 1. Objectives In order to be in a position to conduct a comprehensive research in this paper, we are going to explore the development of hyper spectral remote sensing technology from the time of its discovery in the 1980s to date when number of technologies have been developed and applied in many fields. This will help us identify the importance of hyper spectral remote sensing technology in scientific research, management, engineering and construction. 2. Overview Hyper spectral remote sensing technology is a can be referred to as a technology that combines imagery and spectroscopy to create a single system in order to create capability to deal with large datasets of remote sensed data. This technology can be applied in a number of areas which include: 1. Agriculture: In this area, it can be accomplished by flying aircrafts mounted with hyper spectral imaging devices which capture data. This data is then analyzed and the results are compared to the spectral signature. Figure 1: Analyzed agricultural area in California 2. Study of the atmosphere: satellite imagery can be used to obtain information about the atmospheric water vapor, cloud properties and the effect of aerosols in the atmosphere. This data is then analyzed by scientists and astronomers to determine the extent of atmospheric pollution and make proper predictions. 3. Geology: Hyper spectral remote sensing has technology been used by geologists to analyze the rocks structures, age and distribution and also to determine the availability of different kinds of minerals within the rock structure. 4. Military: Hyper spectral remote sensing technology has been used by the military for quite a long time for military surveillance of target areas. They are able to make an accurate and detailed analysis of the extent, detect any changes and this helps them to make a proper planning of their activities. 5. Ecology and soil monitoring: This technology has been widely used in ecology in monitoring the leaves chlorophyll in different regions, the leaf water in different seasons, the cellulose contained and the pigments available in leaves cover (Cochrane, 2000). In soil monitoring, hyper spectral technology can be used to study the soil properties and the reflectance detected is used to evaluate the soil quality (Irons, et al, 1989). Other parameters that are used to define the soil fertility can also be evaluated using this technology. This includes: organic carbon, nitrogen composition, cations exchange capacity, PH and minerals composition (Baugardner, et al 1985). This gives crucial information that aids in economic planning. 6. Commercial purposes: This technology can be widely used for commercial purposes in areas such as forestry, marine fishing and mineral exploration. Discussion 1. Data collection and processing A hyper spectral imager has a sensor which is mainly composed of some very key components: a) The lens that is used to focus the incoming light. b) A slit which is used to ensure that the incoming light stream is limited to a thin beam. c) A diffraction grating that dispersed the incoming beam into its spectra. d) Photo receptors that have the task of collecting light in specific assigned ranges of the wavelength and then convert the intensities of the band to electrical signals. Figure 2: The components of a hyper spectral remote sensing imager 2. How the Hyperspectral imager works As the atmospheric sunlight illuminates the objects on the ground, a satellite equipped with the hyper spectral imaging sensor may fly over the same area. If such a case occurs, the reflected light is captured and collected by the hyper spectral imaging sensor lens which then focuses it. This light that is collected by the imaging sensor lens then passes through a set of lenses contained within the imaging device. The lenses then focus the light to form an image of the ground information captured. After passing through the lenses, the light enters; the light enters a slit that only allows an extremely thin flat beam to pass. This very thin flat beam is then projected onto a diffraction grating and it is spread into its spectra by the Change Coupled Device (CCD) which is used to measure the spectral intensities of the light that is dispersed. This device converts these measured intensities to electrical signals which are then output for each image pixel. It is important to note that the hyper spectral imaging system has an on board computer which is used to record the output of the Change Coupled Device (CCD) signal. The CCD has a frame rate of 60Hz and every frame captured contains the spectral data for a ground swath that is 0.1m long and 500 km wide. Figure 3: illustration of CCD output 3. Data processing Hyper spectral imaging data is usually collected and represented as a data cube the x-y plane and the z- direction. A spectral cube can be described as an array that has 3 dimensions. It contains spatial data on its x and y axes while the z axes stores spectral information. The spectral cubes usually display information showing difference in stratification, thickness in a three dimensional manner and it thus becomes an important tool for analysis. The extent of the spectral signature can be created by establishing a correlation between the color and the intensity of the spectral response. The magnitude of the spectral signature is created by mapping a color to the intensity of the spectral response at different wavelengths at a given spatial area. This method is dynamic and can be of great help in the visualization of data by mapping out the spectral and spatial features that are not displayable in a single format. The amount of data required to generate hyper spectral data cubes increases as the spectral and spatial dimensions decrease. Image compression is applied to minimize data loss. Figure 4: Illustration of a hyper spectral cube. Figure 5: The concept of hyper spectral imaging. 4. Application of hyper spectral remote sensing technology in management of traffic flow and road conditions. According to Clarke (1999), there have been great advances in hyper spectral remote sensing technology which have demonstrated a lot of capability in analysis of the physical and chemical properties of materials and produce a comprehensive level of results. According to Brecher et al, 2004, there has been a development on the mapping efforts to a certain degree; they have the potential to support transportation infrastructure surveys For civil engineers to assess the conditions of roads and thus have with them the data necessary for proper physical planning there is a common practice among experts whereby they make use of hyper spectral technology in making extensive field observations. They are therefore able to characterize the pavement conditions index (PCI) based on the physical parameters which are established in the assessment. Some of these parameters include: cracking of the roads surface, rutting and raveling. Other technologies that incorporate GIS into hyper spectral Imaging have been developed and they are incorporated into the analysis system in order to determine theses changes. An example of such technology is the pavement management system (PMS). A detailed explanation of the use of hyper spectral technology in the management of traffic flow and road conditions will be explained in details in the section below. A hyper spectral image of the area can be obtained. Today the satellite images occurs from different sources. Figure 6: AVIRIS image of Fairview area with distribution of roads . Road distress surveys must be conducted as a preliminary to the process of planning and design of pavements during rehabilitation. They usually give comprehensive information on the various types of distress present on the roads, location data, and the severity extent of the distresses (Miller and Bellinger, 2002). They also evaluate the pavement conditions taking into consideration the different types of distresses and they incorporate the information obtained into a pavement condition index (PCI). The pavement conditions index (PCI) has a scale 0-100 and after a survey has been conducted, the road is graded using a PCI value. After the road condition has been evaluated, it is the put in categories as the experts may find necessary e.g. excellent, good, fair, poor and very poor. In the figure below, the PCI index for the road surveyed was 76 and it was categorized as Very good. Figure 7: A road condition survey sheet. After the grading and categorization, the next step involves the use of the micro paver pavement management system (PMS) in decision making for us to ensure that we avail a repair that is cost effective for all roads within the area. The pavement management system (PMS) needs to have an additional source of data such as the local county database which will provide digital data for the area. This data is then linked to a mapping programme such as Arc view or the latest production by ESRI ArcGIS which will be used to map and produce a digital road network of the area. The pavement management system (PMS) helps to evaluate the road conditions that currently exist and those likely to occur in future, evaluate the history of the maintenance work and detect any changes that may have occurred. Another surveying technique known as Automatic road analyzer (ARAN) can be used to provide detailed information about road distresses. This system is mounted on a vehicle that has been specially set aside for this survey work. The vehicle that houses this equipment has a large number of sets of computerized devices and sensors which include: laser machines, inertial measurements units, accelerometers, ultrasonic transducers and digital cameras. All the road conditions determined are then integrated into a GIS database. In order to produce an accurate and detailed analysis of the road conditions, we can perform a spectral analysis on the hyper spectral images of the area. This kind of statistical analysis is capable of providing geographically referenced road distresses information for more than 30 parameters. The results obtained are statistically analyzed by analysis of variance (ANOVA) which is generally an F-test that evaluates the compatibility between the distresses identified and the remote sensed data. The statistical analyses allow comparison of the accuracy of the remote sensed signal and the road conditions identified using the PCI. Figure 8: comparison of road survey. Using the statistical results generated from the ANOVA, it is possible to do a comparison and using the spectral characteristics obtained from spectral analysis we can determine the age and type of distress. According to Herold et al, (2004) comparison of the various components of a road surface is possible by analyzing the surface components such as asphalt and bitumen. Asphalt is known to age naturally and this occurs when it reacts with the atmospheric oxygen, the reactions with solar radiation which occur during the day and the effect of the atmospheric heat. According to Bell, 1989, the results can be loss of the oily asphalt components by the effects of volatility; oxidation may change the composition and molecular structure of the surface. These results can be represented in spectra. Figure 9: spectral effects of asphalts ageing. Spectral characteristics of specific damages can be represented using brightness values both old and new (Herold et al, 2004). After the effects of asphalt ageing have been determined, it is finally possible to use the spectral components available to determine the effects of each physical distress parameter. Figure 10: spectral effects of each distress parameter. Figure 11: Results of the ANOVA analysis From the results, an R-squared value of 0.46 and an F-statistic value of 11.24 are a proof of a strong correlation between the remote sensed data and the determined distress parameters. Using the spectral brightness values obtained from hyper spectral images, engineers can determine the extent of traffic jams on the streets. Using hyper spectral images obtained at different interval of time, the spectral brightness characteristics can be mapped and the brightness values obtained can be analyzed using the ANOVA technique. The results obtained can then be used to determine the extent and duration of the traffic jams hence aiding civil engineers to develop more effective traffic management techniques and transport planning. 5. Application of hyper spectral remote sensing technology in transport of fluids through the Pipelines Hyper spectral remote sensing technology can be used to detect pipelines that transport natural resources such as natural gas, carbon (iv) oxide, water and petroleum. However, resources such as petroleum are normally prone to leakages and if their extent is not monitored, the results can be very fatal. Hyper spectral remote sensing can therefore be employed in monitoring the activities in these pipelines. Utility companies can obtain valuable information from digitally captured orthorectified images which they can use for planning, implementation and support of disaster efforts (Herold et al, 2004). A pipeline monitoring system can be used to do track any flaws in the transport system thus ensuring that the proper action is taken promptly when need arises. Figure 11: a pipeline monitoring system In this regard, some of the possible events that may prompt the use hyperspectral remote sensing technology in a pipeline monitoring system include: a) Detection of any effect on the vegetation where leakage has occurred. b) Quantifying the pollution extent where leakage has occurred. c) Monitoring the natural gas, after proper remedies have been taken. A pipeline monitoring system ensures that the transport lines are kept under surveillance at all times and incase of any leakage, it can be recognized and stopped before its effects become catastrophic. The process of monitoring the pipelines entails: 1. A satellite monitors all the events occurring at the pipelines area and it may detect an unusual event which it records. 2. Over a period of time, the satellite captures several images of the same area which are then transmitted to the ground based station. 3. The analyst on the ground analyzes the images and detects change in the images relayed. 4. Satellite detected encroachment is identified Figure 12: satellite detected encroachment. 5. T he encroachment is geographical referenced using the available tools such as ERDAS Imagine or any other appropriate program. Figure 13: supervised classified image of the leakage area Finally, an alert is sent to the field for proper measures to be taken and avert any possible disaster. However, while the proper measures are being taken, the pipeline monitoring system monitors the spread of the leaked fluid and relays images at different intervals of time which are analyzed and that way the authorities can assess their progress. Finally, the monitoring system is used to detect any changes in the environment that may have resulted from the leakage and this helps in advising the inhabitants of the area on the proper health measures to take. Summary This research has extensively explored the field of hyper spectral remote sensing from the time it was developed in the 1980s to the current technologies in this field. From the discussions, the importance of hyper spectral remote sensing technology cannot be overlooked as it is applicable in a wide number of areas particularly in civil engineering. Structures such as roads which are designed to last for quite a considerable period of time are monitored using this technology and any change that may occur can be noted and the necessary measures taken to rectify it and avoid any further damage. It can also be employed in urban areas that experience traffic jams and proper management policies can be enacted. From the discussion, hyper spectral remote sensing technology is also very important in monitoring the transport of fluids through pipelines so that any case of leakage can be reported and the necessary steps taken to avoid any disaster. As research continues in this field, there is a likelihood of more efficient algorithms being invented thus making hyper spectral remote sensing applicable in more fields. References Dutta, S., and Alam, (1995), W., Oil & Gas Exploration and Production Waste – RCRA Exemptions and Non-exempts, National Conference, Society of Petroleum Engineers, Houston, Texas, Yingchun. S, Z. Xianfeng, C. Xiuwan, H. Zhaoqiang & W. Caicong (2006). Mangrove type classification using airborne hyperspectral images at Futian reservation, Shenzhen, China. Geoscience and Remote Sensing Symposium 2006, IGARSS 2006, IEEE International Conference; pp 3451-3454 Tong,Q.,Zhang,B.,&Zheng,L.(2004).Hyperspectral Remote Sensing Technology and Applications in China. Proc. Of the 2nd CHRIS/Proba Workshop, Frascati, Italy. Clark, R.N., 1999. Spectroscopy of Rocks and Minerals and Principles of Spectroscopy, In: A.N. Rencz (ed.).Manual of Remote Sensing, Chapter 1, John Wiley and Sons, New York, pp. 3-58. Asner. G.P, D.E Knapp, T. Kennedy-Bowdoin, M.O Jones, R.E Martin, J. Boardman & C.B Field (2007). Carnegie Airborne Observatory: in-flight fusion of hyperspectral imaging and waveform LiDAR for 3D studies of ecosystems. Journal of Applied Remote Sensing. Vol 1; pp 1-27 Boschetti. M, L. Boschetti, S. Oliveri, L. Casati & I. Canova (2007). Tree species mapping with airborne hyperspectral MIVIS data: the Ticino Park case study. International Journal of Remote Sensing. Vol 28; Issue 6; pp 1251-1261 Read More