Remote Sensing and Image Processing - Essay Example

? Report - Remote Sensing and Image Processing Remote Sensing and Image Processing Introduction This report demonstrates the semantics related to events that were accomplished in practical classes associated with the domain of remote sensing and image processing. Likewise, the events were incorporated with estimation of vegetation shield by consuming two indexes i.e. using Normalized Difference Vegetation Index (NDVI) and Ratio Vegetation Index (RVI). Moreover, along with these indexes, supervised and unsupervised classification is also conducted that is related to different information classes illustrated on an image. Moreover, these events also include images and pictures that are captured via satellites that are powered with remote sensing technology. The advantage of remote sensing is that it can detect objects in real time along with providing coverage of an expanded area within a short period of time. However, remote sensing can be divided in to two types i.e. active remote sensing and passive remote sensing. The active remote sensing provides opportunity to be utilized at any time throughout the day and regardless of any season. Moreover, active remote sensing also facilitates the inspection of wavelengths that are not effectively illustrated by the Sun. In addition, it also takes charge of the illumination on the targeted object. On the other hand, passive remote sensing requires the Sun, in order to brighten the objective or the target. Moreover, this type of remote sensing uses reflected waves to measure the distance. As (Smith 2010) states that remote sensing based on active methodology needs a significant amount of energy source for illuminating the target. The data for the process of task that was associated with the vegetation estimation exposure was gathered from the National Oceanic and Atmospheric Administration's (NOAA) Advanced Very High Resolution Radiometer (AVHRR). In order to achieve a detection of vegetation, band rationing of the value of Infra-Red (700-1300nm) by Visible Light (400-700nm) is utilized. By using these values and methods, the albedo effects will be eliminated along with issues related to shadows that emerge from the images that are processed. Moreover, this method will also facilitate high quality visibility for vegetation in images. Moreover, a task that is related to classification of information classes, Churn Farm image is the best option. Likewise, data available in this image is gathered by an airplane in the year 1984 from the NERC ATM scanner. Likewise, the image includes four bands and sketches the sites that are associated with agricultural land use. Moreover, integer represents the cover type for specific land cover type, in this way; probable training sites can be marked for each type of land cover. Task 1 - Vegetation Index Methodology This task utilized a methodology including calculation of Ratio Vegetation Index (RVI) and Normalized Difference Vegetation Index (NDVI). The NDVI will be used for the image of Lake Nakuru Thermatic Mapper (TM). Likewise, for data associated with TM, two bands are utilized i.e. band 3 and band 4. Band 3 calculates the red light and band four calculates the red infra light. Outside the scope of the town named as Nakuru, a small salt ware lake named as ‘lake Nakuru’ is located. Likewise, the lake is famed for the spectacle, as approximately one million flamingos comes to it for feeding themselves with green algae located in the warm water areas of the lake. Moreover, pelicans often come here to feed themselves with cormorants that are also available deep in the lake (Smith 2010). Steps Implemented Steps are demonstrated in points below: Considering as a Raster Layer, open the image file named as nakuru.img. In the available options, select the option to view the image with channel 4 that will be associated with green gun and red gun. However, channel 2 will demonstrate blue gun. From the main menu options tab, select the option ‘Image Interpreter’ and select and click the option named as ‘Spectral Enhancement’ After reaching the menu associated with ‘Spectral Enhancement’, execute the ‘Indices’ option to compute indices. Next step is to select the input image file named as nukuru.img and select output image file named as nakuru_rvi.img. Always crosscheck that the output options always highlights Landsat TM as the output sensor. Moreover, configure the IR/R options by utilizing the ‘select function box’ as this will facilitate the parameters of the output image to be set as a direct ration for Channel 3 and Channel 4. Lastly, click ‘OK’ to start process. After the completion of the process, execute the Raster Attribute Editor to modify start and end colors for the image file named as nakuru_rvi.img Computing Normalized Difference Vegetation Index (NDVI) Computing In order to compute NDVI, open the image file named as ‘nakuru.img’. This file will be executed as a Rester Layer. Likewise, the next step will be to set the green guns and red guns with channel 4 and blue guns with channel 2 Now execute the dialog box of the interpreter that is located at the main menu. After executing the interpreter, select option associated with ‘Spectral Enhancement’ The next step will be to click on the options associated with indices for calculating indices. Configure the options for the image named as nakuru.img as Input file and nakuru_ndvi.img as the Output file. Check the configuration setting as the requirement is to set the sensor in the output option as NDVI. After configuring output options, IR/R options will be set from the ‘select function box’. Click OK in order to start process After the completion of the process, once again execute the Raster Attribute Editor for modifying start colors and end colors for the image file named as nakuru_ndvi.img. After completing thee steps, associate these three images with each other i.e. nakuru.img, nakuru_rvi and nakuru_ndvi.img Advantages of using Ratio Vegetation Index (RVI) and Normalized Difference Vegetation Index (NDVI) In order to estimate vegetation by deploying RVI and NDVI, there are various advantages and some disadvantages too. We know that RVI computes via a spectral ratio and it is not affected by scene illumination. Moreover, by incorporating spectral ratio, it can be distinguished that there are spectral differences. Conversely, some of the negative factors associated with RVI includes accentuates noise, this type of noise must be eliminated before rationing begins. In comparison, NDVI is recommended due to the fact that if facilitates compensation for illuminations, slope and aspect that are modifying. Moreover, it also demonstrates precise results due to the fact of calculating two bands. Although, it is profound in terms of incident solar radiation, which is used to gather data, atmospheric conditions and off-nadir viewing. Results Achieved If we compare the two images, we can conclude that NDVI demonstrates a precise and perkier image of the vegetation as shown in the original image. Below is the illustration of figure 1, 2 and 3 correspondingly with RVI and NDVI processed images. Figure 1 Original Image - Nakuru Lake (Provided) Figure 2 Image Produced by RVI processing Figure 3 Image Produced by NDVI Processing Discussion In the discussing heading, indices associated with RVI and NDVI were calculated of an image associated with Nakuru Lake. Above that are demonstrated above, one can conclude that NDVI shows clarity and precision in terms of accuracy in terms of description of vegetation indices in comparison to images that are based on RVI. The reason for this significant different between these two includes accurate technical mechanism i.e. normalized 2-band. Likewise, this band is used for calculation of indices that facilitates the factor of compensation for angle of look variables, illumination and scaling between limits that are known. Task 2 - Unsupervised and Supervised Classification Methodology The methodology for this course work utilized both the supervised and the unsupervised methodologies for conducting image classifications. The unsupervised methodology is injected in ‘K means’. Likewise, in this process, dissimilar kinds of information classes are used from the images, as for this purpose, initial means are nominated. Besides, the outcomes from these steps are utilized to construct boundaries via dissimilar parts possessing various information classes. On the other hand, the supervised method is implemented by the association of these three elements i.e. "Maximum Likelihood", "Minimum Distance to Means" and "Parallelpiped". Likewise, this methodology includes class samples that are incorporated for training the software, as the software has to perform the task of classification on the images that will be provided (Smith 2010). Steps Performed Unsupervised Classification 1) Open the provided image file as a Raster Layer and set the Layers to Colors 1(Red), 2(Green) and 3(Blue). This file is shown in Figure 4. Figure 4 Original Image - churn_all.img 2) Open up the Signature Editor from the Classification Menu of the ERDAS software. 3) Use the AOI Tools to collect signature for different information classes. a) Use the Polygon Tool from the AOI Toolbox to create a polygon on a wheat field. b) Add this AOI to the Signature Editor window by clicking on the Add AOI icon. c) Change the Signature Name to Wheat Field_1. d) Change the color of the Signature to Blue. e) Repeat steps a,b,c and d on multiple wheat fields to create variance in the data of the information class. f) Merge all signature of the Wheat information class by selecting the signature merge button. Delete all other classes except this one and rename it to "Wheat". 4) Repeat step 3 for all of the remaining classes in the image. 5) Using the created signatures, perform K-means classification on the image. 6) Use the Histogram tool to evaluate the signatures created using the above method. Supervised Classification The supervised classification methods can be used through the Supervised Classification dialog box of the ERDAS software. For Maximum Likelihood Classifier 1) Open the Supervised classification dialog box and set Input Raster File to churn_all.img, Input Signature File to churn.sig and Output File to churn_max.img. 2) Set Non-Paratmetric Rule to None. 3) Set Parametric Rule to Maximum Likelihood. 4) Click OK to run the classifier 5) After the classification is done, open the churn_max.img file and select Pseudo Color from the Raster Option tab. 6) Compare this file with the churn_ref.img file which shows how the classification of the image should look like. For Minimum Distance to Mean 1) Repeat the steps given for Maximum Likelihood Classifier but set the Output File to churn_mind.img and Parametric Rule to Minimum Distance. 2) Run the classifier and compare with the churn_ref.img file. For Parallelpiped 1) Repeat the steps given for Maximum Likelihood Classifier but set the Output File to churn_piped.img and Parametric Rule to Parallelpiped. Also, set both Overlap Rule and Unclassified Rule to Unclassified. 2) Run the classifier and compare with the churn_ref.img file. Strength and Weaknesses of Unsupervised Classification There are many advantages by using the unsupervised classification methodology. Likewise, the advantages includes the factor of no requirements of the region related to past information, little risk of generating human errors and acknowledgement of various classes representing as distinct units. Conversely, there are many disadvantages of using this approach, as unsupervised classification methodology strongly relies on natural grouping associated with data. Issues related to matching these different groups with the filed data can be expected. Likewise, the issues include dissimilarity among information classes and spectral, analyst controls the image classes inappropriately and frequent changes of spectral properties. Strength and Weaknesses of Supervised Classification By incorporating the supervised classification methodology, there are advantage including excessive control for an analyst, there are no mandatory requirements for identical spectral classes with information classes, only known areas are processed. Although, by following this methodology, there is a factor that enforces structure on the data. Moreover, the nominated areas of training will not demonstrate the conditions on the complete image. Furthermore, it is not possible to point out the classes that are absent in the training data. Results Achieved Figure 5 demonstrates the output image associated with unsupervised classification and figures 5, 6, and 7 displays the histograms that is constructed for dissimilar signatures. Figure 5 Unsupervised Classification Figure 6 Histogram for Lucerne Figure 7 Histogram for Barley Figure 8 Histogram for Wheat Figure 8, 9 and 10 express the results accomplished from three types of supervised classification completed in Task 2 that are Maximum Likelihood, Minimum Distance to Mean and Parallelpiped. Discussion Under the discussion heading, classification was processes for different information classes associated with the image of Churn Farm. Likewise, both of the methods of classification namely, unsupervised and supervised were implemented. In the unsupervised classification methodology, there is no requirement of past information of the region. Moreover, the methodology also facilitates distinct recognition of dissimilar information classes. Conversely, it does provide an efficient control to the analyst for the image. Although, the training data constraints various distinct classification that can be conducted. Likewise, by associating the three images that are constructed by different methods of supervised classification, the difference is viable, as maximum likelihood classifier demonstrates a result that is leaning towards originality of the reference image, as compared to the other two methods. Conclusion Incorporating remote sensing facilitates correct and precise surveillance of objects in a real time environment and without wires. Likewise, it also assist the observation of bands that is not effectively provided by the Sun. The report demonstrates two tasks i.e. vegetation coverage of Nukura Lake and classification of information classes. The first task, estimation of Nukru Lake was conducted by deploying vegetation indices. The task was facilitated by two core factors i.e. Ratio Vegetation Index (RVI) and Normalized Difference Vegetation Index (NDVI). The observations concluded that NDVI is more efficient, as compare to RVI. The second task was associated with both supervised and unsupervised methodologies that were performed for the classification of information classes that were available in the image named as Churn Farm. The unsupervised method includes a construction of signatures belonging from dissimilar information. Conversely, supervised methodology includes training software with the aid of data that is used for software classification. By associating all the three results from supervised classification methodologies implemented in task 2, the conclusion states that the maximum likelihood classified methodology constructed accurate results, as compared to the other two methods. References Smith M. 2010. Lecture Notes. Remote Sensing and Image Processing. Kingston University. Read More