Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
SlideShare a Scribd company logo

IRJET - Normalized Difference Vegetation Index (NDVI) based Land Cover Classification using ArcGIS

•
•
IRJET Journal
IRJET Journal

This document discusses using ArcGIS and remote sensing to classify land cover in Vijayawada, India using NDVI values derived from Landsat8 imagery. Landsat8 images from different dates were analyzed to calculate NDVI and classify the land cover into vegetation, water bodies, bare soil, and urban areas. Supervised classification was performed and accuracy assessment using an error matrix showed overall accuracy of 88% and kappa coefficient of 0.85. The study aims to identify land cover changes over time and provide information for urban planning and management.

Related slideshows

Alexander vega 2019_iop_conf._ser.__mater._sci._eng._603_022010
Alexander vega 2019_iop_conf._ser.__mater._sci._eng._603_022010Alexander vega 2019_iop_conf._ser.__mater._sci._eng._603_022010
Alexander vega 2019_iop_conf._ser.__mater._sci._eng._603_022010
 
30. z. t. khan and dipankar bera
30. z. t. khan and dipankar bera30. z. t. khan and dipankar bera
30. z. t. khan and dipankar bera
 
12 ijcse-01224
12 ijcse-0122412 ijcse-01224
12 ijcse-01224
 
1 of 6
Download to read offline
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2895 Classification using ArcGIS A.V. Kiranmai1, N. Aparna2, B. Jyothi Sree3, D. Pavan Kalyan4, K. Suresh5 1 Assoc. Professor, Departments of Electronics and Communication Engineering, Potti Sriramulu Chalavadi Mallikarjuna Rao College of Engineering and Technology, Vijayawada, India. 2,3,4,5 student, Departments of Electronics and Communication Engineering, Potti Sriramulu Chalavadi Mallikarjuna Rao College of Engineering and Technology, Vijayawada, India. --------------------------------------------------------------------***-------------------------------------------------------------------------- Abstract—The Normalized Difference Vegetation Index(NDVI) is one of the most widely used vegetation index which uses red(RED) and near infrared (NIR)bands of electromagnetic spectrum, and is utilized for the analysis of remote sensing images to obtain the vegetation information of the target area. In this paper the land cover classification of Vijayawada region, Krishna district, Andhra Pradesh is analyzed using NDVI.Landsat8 images of different dates were collected and derived NDVI values. Versatile bands of Landsat images are used to acquire the information of vegetation, water bodies, bare soil, and urban by calculating NDVI. Present study concentrated on making out the difference between the vegetation indexes of various land cover types by performing supervised classification. Key Words: Remote sensing, Normalized Difference Vegetation Index (NDVI), Land cover, multi spectral images. 1. INTRODUCTION Earth environment can be understood in a better way using Remote Sensing technology [1]. It is the Science and Art of getting information and extracting the attributes in form of Spectral, Spatial and Temporal about few physical objects, area development, such as vegetation, land cover classification, urban area, agriculture land and water resources without coming into physical contact of these objects [2]. Landsat is one of the most used remote sensing satellites .Features of Landsat 8 are as follows: It has two different operating sensors i) Operational Land Imager (OLI) ii) Thermal Infrared Sensor (TIRS) sensors, Which have a total of 11 bands, consisting of 9 bands (band 1 - 9) situated at OLI and 2 bands (bands 10 and 11) at TIRS [3]. Change detection of Land cover is one of the most important techniques, widely utilized for planning and managing land [4]. In recent years, geographic information system and remote sensing have been hired for several applications including land cover change detection [5]. Vegetation coverage for various time periods and for peculiar areas can be obtained using Normalized Difference Vegetation Index(NDVI) .The NDVI aims to find land cover change caused by human action such as construction and development, and to analyze changes in vegetation because of environmental changes [6]. The NDVI is used as an index of measurement of balance among the energy accepted and gives out by earth objects [7]. The NDVI can be computed by the combination of Red and NIR bands of Landsat8 images. This paper mainly aims to find out land cover changes of Vijayawada city by using the NDVI. REDNIR REDNIR NDVI    Normalized Difference Vegetation Index (NDVI) based Land Cover
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2896 2. STUDY AREA Vijayawada is a historic city located at the geographical centre of Andhra Pradesh state in India on the banks of Krishna River with latitude 16003’11” N and longitude 800 03’91” E. The climate is equatorial, with hot summers and moderate winters. The peak temperature reaches 47 °C in May-June, while the winter temperature is 20-270 C. The mean humidity is 78% and the mean annual rainfall is 103 cm.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2897 3. METHODOLOGY: Fig. 1: Methodology This study utilized Landsat satellite images of 2019. Landsat OLI images, each with a 30 m resolution, are gathered from United States Geological Survey (USGS) website, Earth Explorer (http://earthexplorer.usgs.gov/).Band 4 and Band5 are used for land cover classification of ROI where ROI in this study is Vijayawada. The band description of Landsat8 images are shown below
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2898 TABLE 1. Characteristics of OLI and TIRS sensors on image Landsat 8 TABLE 2. Band characteristics for object imagery on image Landsat 8 Band Description 1 Aerosol studies and coastal areas 2 Bathymetric mapping, differentiates soil from vegetation and leaves from conifer vegetation 3 Emphasize the peak of vegetation to assess the strength of vegetation 4 Differentiate the vegetation angle 5 Emphasize the content of biomass and coastline 6 Discriminates soil water content and vegetation; penetrate a thin cloud 7 Enhancement soil moisture content and vegetation, thin cloud penetration 8 Resolution 15 m, image sharpening 9 Enhancement detection of contaminated sirus clouds 10 Resolution 100 m, temperature mapping and soil moisture counting 11 Resolution 100 m, enhancement of temperature mapping and soil moisture counting Input Landsat8 OLI images are acquired from United States Geological Survey (USGS). Band4 and Band5 are combined using composite bands. NDVI technique is applied on composite signal generated. Using Interactive supervised classification image is classified into various types like agriculture, vegetation, urban barren land, and water. Classified data is validated with ground truth data and error matrix is generated. Quality parameters (Overall Accuracy, kappa coefficient) are calculated using error matrix. 4. RESULTS AND DISCUSSION: In this work Landsat8 images are acquired using USGS database and classified using NDVI to obtain overall accuracy. The acquired image is of Vijayawada area, Andhrapradesh, India. Landsat images are analyzed using Remote Sensing and ArcGIS 10.3 software. The classified image contains five classes built-up, agriculture, vegetation, scrub land and water bodies. The error matrix is calculated and obtained results are compared with previous results. Table 3: Class wise area for years 2008, 2014, 2019 Year 2008(Area in hectares) 2014(Area in hectares) 2019(Area in hectares) Built-up 6370.24 6440.20 6905.25 Agriculture 20791.00 19248.20 18358.86 vegetation 3654.01 5908.34 6253.05 Scrub land 1886.97 2061.09 2210.03 Water bodies 1108.12 962.51 834.06
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2899 Fig. 2: Land cover change Table 4: Error matrix of classified image Land Cover Built-up Agriculture vegetation Scrub land Water bodies Column total Built-up 13 1 0 1 0 15 Agriculture 0 13 0 1 1 15 vegetation 1 0 12 0 0 13 Scrub land 0 0 1 12 0 13 Water bodies 0 1 1 0 12 14 Row total 13 15 15 14 13 70 From the above table (using Error matrix) the Overall Accuracy (OA) and kappa coefficient are obtained as 88%, 0.85. 5. CONCLUSION: The NDVI is an important indicator used to determine land use and land cover changes because of urban and economic development. Changes in land cover can be observed through extracted values of the NDVI map. From the results it is observed an amount of decrement in dense vegetation due to various reasons such as urban and economic development because of population growth and deforestation. Results obtained can be used as indicators for future trends to get land cover changes and for distinguishing effective factors on vegetation cover for the improved understanding of planners and decision makers on the issue. 6. REFERENCES: 1. Ahmadi H, Nusrath A, “Vegetation change Detection of Neka river in Iran by using remote sensing and GIS”, Journal of geography and Geology, 2 (1)., pp. 58-67., 2012. 2. Karaburun A. A. K. Bhandari, “Estimation of C factor for soil erosion modelling using NDVI in Buyukcekmece watershed”, Ozean Journal of applied sciences 3, 77-85., 2010. 3. National Institute of Aeronautics and Space, “Guidelines for Utilization of Landsat-8 Data for Detection of Inundated Area,” (2015). 4. Sahebjalal, E. and K. Dashtekian, Analysis of land use-land covers changes using normalized difference vegetation index (NDVI) differencing and classification methods. African Journal of Agricultural Research, 2013. 8(37): p. 4614- 4622.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2900 5. Maasikamäe, S., H. Hass, and E. Jürgenson. The Impact of Uncontrolled Development on the Use of Arable Land. in Proceedings: The Fifth International Scientific Conference Rural Development 2011, 24-25 November, 2011, Akademija. 2011. 6. Nath, B., Quantitative Assessment of Forest Cover Change of a Part of Bandarban Hill Tracts Using NDVI Techniques. Journal of Geosciences and Geomatics, 2014. 2(1): p. 21-27. 7. Meneses-Tovar, C., NDVI as indicator of degradation. Unasylva (FAO), 2012.

More Related Content

IRJET - Normalized Difference Vegetation Index (NDVI) based Land Cover Classification using ArcGIS

  • 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2895 Classification using ArcGIS A.V. Kiranmai1, N. Aparna2, B. Jyothi Sree3, D. Pavan Kalyan4, K. Suresh5 1 Assoc. Professor, Departments of Electronics and Communication Engineering, Potti Sriramulu Chalavadi Mallikarjuna Rao College of Engineering and Technology, Vijayawada, India. 2,3,4,5 student, Departments of Electronics and Communication Engineering, Potti Sriramulu Chalavadi Mallikarjuna Rao College of Engineering and Technology, Vijayawada, India. --------------------------------------------------------------------***-------------------------------------------------------------------------- Abstract—The Normalized Difference Vegetation Index(NDVI) is one of the most widely used vegetation index which uses red(RED) and near infrared (NIR)bands of electromagnetic spectrum, and is utilized for the analysis of remote sensing images to obtain the vegetation information of the target area. In this paper the land cover classification of Vijayawada region, Krishna district, Andhra Pradesh is analyzed using NDVI.Landsat8 images of different dates were collected and derived NDVI values. Versatile bands of Landsat images are used to acquire the information of vegetation, water bodies, bare soil, and urban by calculating NDVI. Present study concentrated on making out the difference between the vegetation indexes of various land cover types by performing supervised classification. Key Words: Remote sensing, Normalized Difference Vegetation Index (NDVI), Land cover, multi spectral images. 1. INTRODUCTION Earth environment can be understood in a better way using Remote Sensing technology [1]. It is the Science and Art of getting information and extracting the attributes in form of Spectral, Spatial and Temporal about few physical objects, area development, such as vegetation, land cover classification, urban area, agriculture land and water resources without coming into physical contact of these objects [2]. Landsat is one of the most used remote sensing satellites .Features of Landsat 8 are as follows: It has two different operating sensors i) Operational Land Imager (OLI) ii) Thermal Infrared Sensor (TIRS) sensors, Which have a total of 11 bands, consisting of 9 bands (band 1 - 9) situated at OLI and 2 bands (bands 10 and 11) at TIRS [3]. Change detection of Land cover is one of the most important techniques, widely utilized for planning and managing land [4]. In recent years, geographic information system and remote sensing have been hired for several applications including land cover change detection [5]. Vegetation coverage for various time periods and for peculiar areas can be obtained using Normalized Difference Vegetation Index(NDVI) .The NDVI aims to find land cover change caused by human action such as construction and development, and to analyze changes in vegetation because of environmental changes [6]. The NDVI is used as an index of measurement of balance among the energy accepted and gives out by earth objects [7]. The NDVI can be computed by the combination of Red and NIR bands of Landsat8 images. This paper mainly aims to find out land cover changes of Vijayawada city by using the NDVI. REDNIR REDNIR NDVI    Normalized Difference Vegetation Index (NDVI) based Land Cover
  • 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2896 2. STUDY AREA Vijayawada is a historic city located at the geographical centre of Andhra Pradesh state in India on the banks of Krishna River with latitude 16003’11” N and longitude 800 03’91” E. The climate is equatorial, with hot summers and moderate winters. The peak temperature reaches 47 °C in May-June, while the winter temperature is 20-270 C. The mean humidity is 78% and the mean annual rainfall is 103 cm.
  • 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2897 3. METHODOLOGY: Fig. 1: Methodology This study utilized Landsat satellite images of 2019. Landsat OLI images, each with a 30 m resolution, are gathered from United States Geological Survey (USGS) website, Earth Explorer (http://earthexplorer.usgs.gov/).Band 4 and Band5 are used for land cover classification of ROI where ROI in this study is Vijayawada. The band description of Landsat8 images are shown below
  • 4. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2898 TABLE 1. Characteristics of OLI and TIRS sensors on image Landsat 8 TABLE 2. Band characteristics for object imagery on image Landsat 8 Band Description 1 Aerosol studies and coastal areas 2 Bathymetric mapping, differentiates soil from vegetation and leaves from conifer vegetation 3 Emphasize the peak of vegetation to assess the strength of vegetation 4 Differentiate the vegetation angle 5 Emphasize the content of biomass and coastline 6 Discriminates soil water content and vegetation; penetrate a thin cloud 7 Enhancement soil moisture content and vegetation, thin cloud penetration 8 Resolution 15 m, image sharpening 9 Enhancement detection of contaminated sirus clouds 10 Resolution 100 m, temperature mapping and soil moisture counting 11 Resolution 100 m, enhancement of temperature mapping and soil moisture counting Input Landsat8 OLI images are acquired from United States Geological Survey (USGS). Band4 and Band5 are combined using composite bands. NDVI technique is applied on composite signal generated. Using Interactive supervised classification image is classified into various types like agriculture, vegetation, urban barren land, and water. Classified data is validated with ground truth data and error matrix is generated. Quality parameters (Overall Accuracy, kappa coefficient) are calculated using error matrix. 4. RESULTS AND DISCUSSION: In this work Landsat8 images are acquired using USGS database and classified using NDVI to obtain overall accuracy. The acquired image is of Vijayawada area, Andhrapradesh, India. Landsat images are analyzed using Remote Sensing and ArcGIS 10.3 software. The classified image contains five classes built-up, agriculture, vegetation, scrub land and water bodies. The error matrix is calculated and obtained results are compared with previous results. Table 3: Class wise area for years 2008, 2014, 2019 Year 2008(Area in hectares) 2014(Area in hectares) 2019(Area in hectares) Built-up 6370.24 6440.20 6905.25 Agriculture 20791.00 19248.20 18358.86 vegetation 3654.01 5908.34 6253.05 Scrub land 1886.97 2061.09 2210.03 Water bodies 1108.12 962.51 834.06
  • 5. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2899 Fig. 2: Land cover change Table 4: Error matrix of classified image Land Cover Built-up Agriculture vegetation Scrub land Water bodies Column total Built-up 13 1 0 1 0 15 Agriculture 0 13 0 1 1 15 vegetation 1 0 12 0 0 13 Scrub land 0 0 1 12 0 13 Water bodies 0 1 1 0 12 14 Row total 13 15 15 14 13 70 From the above table (using Error matrix) the Overall Accuracy (OA) and kappa coefficient are obtained as 88%, 0.85. 5. CONCLUSION: The NDVI is an important indicator used to determine land use and land cover changes because of urban and economic development. Changes in land cover can be observed through extracted values of the NDVI map. From the results it is observed an amount of decrement in dense vegetation due to various reasons such as urban and economic development because of population growth and deforestation. Results obtained can be used as indicators for future trends to get land cover changes and for distinguishing effective factors on vegetation cover for the improved understanding of planners and decision makers on the issue. 6. REFERENCES: 1. Ahmadi H, Nusrath A, “Vegetation change Detection of Neka river in Iran by using remote sensing and GIS”, Journal of geography and Geology, 2 (1)., pp. 58-67., 2012. 2. Karaburun A. A. K. Bhandari, “Estimation of C factor for soil erosion modelling using NDVI in Buyukcekmece watershed”, Ozean Journal of applied sciences 3, 77-85., 2010. 3. National Institute of Aeronautics and Space, “Guidelines for Utilization of Landsat-8 Data for Detection of Inundated Area,” (2015). 4. Sahebjalal, E. and K. Dashtekian, Analysis of land use-land covers changes using normalized difference vegetation index (NDVI) differencing and classification methods. African Journal of Agricultural Research, 2013. 8(37): p. 4614- 4622.
  • 6. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072 © 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 2900 5. Maasikamäe, S., H. Hass, and E. Jürgenson. The Impact of Uncontrolled Development on the Use of Arable Land. in Proceedings: The Fifth International Scientific Conference Rural Development 2011, 24-25 November, 2011, Akademija. 2011. 6. Nath, B., Quantitative Assessment of Forest Cover Change of a Part of Bandarban Hill Tracts Using NDVI Techniques. Journal of Geosciences and Geomatics, 2014. 2(1): p. 21-27. 7. Meneses-Tovar, C., NDVI as indicator of degradation. Unasylva (FAO), 2012.