Professional Documents
Culture Documents
Land use/land cover classification plays an important role in the sustainable management of the natural
resources. Remote sensing techniques have been intensively used for land use /land cover
classification. Over the last decade new technologies had emerged that increased the accuracy of the
produced maps. The aim of this work is to investigate for an effective image classifier by comparing
three classifiers i.e. Maximum Likelihood classifier (MLC), Artificial Neural Network (ANN) and Support
Vector Machine (SVM) for land use/ land cover classification for a selected area in the north of the Delta
+
of Egypt. In this work, the surface reflectance data of Landsat 7 ETM was used. To reduce the
redundancy in the data, the Principal Component Analysis (PCA) was used and the first three PCAs
were selected for further analysis. The accuracy of the three classification method was compared using
the confusion matrix. In this work the overall accuracy of ANN was satisfactory using both the ENVI's
Neural Net and the neural network Pattern Recognition tool of Matlab and therefore was not
recommended in this study area. The overall accuracy for MLC and SVM was 80.28 and80.64 %,
respectively. Both MLC and SVM showed similar overall accuracy with an advantage of MLC when
considering the producer and user accuracy. To enhance the classification results, the Normalized
Difference Water Index (NDWI) was included and the MLC and SVM were re-examined. This addition
enhanced the overall accuracy of MLC to 83.95 % while the SVM decreased to 79.03%. Thus, it is
recommended using the MLC when classifying the land use/ land cover in this study area using the
same approach.
Keywords: Maximum likelihood classification, artificial neural networks, support vector machine, land use/ land
cover, image classification
INTRODUCTION
Land use/land cover classification is essential in prevent land degradation. Remote sensing is considered a
developing sustainable land management strategies rapid and lower cost method compared to ground surveys.
especially those aiming to improve water quality and Image classification is one of the important parts in remote
sensing applications for the land use and land cover
mapping. However, it is not easy to generate a satisfactory
result because of the limitation of data and complexity of
*Corresponding Author's Email: randa_sgmm@yahoo.com the land use and cover types (Wong and Sarker, 2014).
Morgan et al. 811
The Maximum Likelihood classifier (MLC)is one of the insensitive to dimensionality. SVMs have often been found
most popular classification method that was used for land to provide higher classification accuracies than other
use/ cover classification (Shalaby and Tateishi, 2007; widely used pattern recognition techniques, such as the
Manandhar et al., 2009; Ahmad and Quegan, 2012;and maximum likelihood. Furthermore, SVMs appear to be
Abdikan et al., 2015). MLC is a parametric classifier which especially advantageous in the presence of heterogeneous
assumes normal or near normal spectral distribution for classes for which only few training samples are available
each feature of interest. Jensen (2005) reported that it is (Moughal, 2013). Therefore, SVM is considered one of the
based on the probability that a pixel belongs to a particular most favorable classification techniques (Carrao et al.,
class and takes the variability of classes into account by 2008; Szuster et al., 2011;Thai et al 2012 and Moughal,
using the covariance matrix. Nevertheless, it requires 2013).
sufficient ground truth data and it cannot be applied if the The aim of this work is to investigate for an effective
distribution of the population does not follow the normal image classifier by comparing three classifiers i.e.
distribution (Japan Association of Remote Sensing, 1999). Maximum Likelihood classifier (MLC), Artificial Neural
In real life applications, the nature of the distribution is Network (ANN) and Support Vector Machine (SVM) for
hardly known and therefore, it is preferable to use land use/ land cover classification for a selected area in the
nonparametric classifiers that are free from assumptions north of the Nile Delta, Egypt.
(Moughal, 2013).
Artificial Neural Networks (ANN)are nonlinear analytic
techniques modeled after the hypothesized processes of MATERIALS AND METHODS
learning in the cognitive system and the neurological
functions of the brain and capable of predicting new Study Area
observations from other observations after executing a
process of so-called learning from existing data(Dell Inc., The study area is located in the northern part of the Nile
2013).They offer the potential to classify data based upon Delta, between longitudes 30 32' 52" and 31 8' 31" E and
a rapid match to overall patterns using previously latitudes 31 16' 26" and 31 36' 50" N, and covers an area
2
calculated weighting factors, rather than point by point of about 1650 km with the Mediterranean Sea representing
comparisons involving algorithmic logic applied to its northern border (Map 1).The study area included El-
individual (Miller et al., 1995). Furthermore, Neural Burullus Lake, the sandy wetlands, the old clayey deltaic
networks make no assumption about data distribution (Qiu soils, the relatively new reclaimed soils clayey soils(which
and Jensen, 2004).ANN are applicable in virtually every were dried from the clay flats and swamps that surrounded
situation in which a relationship between the predictor the lake) and the new reclaimed sandy bare soils. The
variables and predicted variables exists, even when that main land use in the study area is agriculture; mostly field
relationship is very complex and not easy to articulate in crops in the north delta area are rice, barely, beans and
the usual terms of "correlations" or "differences between clover. Vegetables including tomatoes, potatoes,
groups" (Dell Inc., 2013).Qiu and Jensen, (2004) reported watermelon and others are cultivated in small scattered
that improved image classification results compared to the areas. Fish ponds are the second dominant land use in the
traditional methods and various studies documented such study area. El-Burullus Lake is a shallow brackish lake
improvements in land use/ land cover classification characterized by a few species of abundant aquatic
(Ashish, 2002;Pacificiet al., 2009; Yuan et al., 2009 and vegetation.
Ndehedehe et al., 2013). The study area is subjected to numerous ecological threats
Support vector machine (SVM) is one of the latest such as drying parts of the lake, the fast growth of fish
developed image classification techniques, it is a ponds mainly around the southern fringes of the lake, lake
supervised classification method derived from statistical water pollution due to the excessive inflow of agriculture
learning theory, which describes the properties of learning drainage water, and the fast growing new human
machines that allow them to give reliable predictions and settlements. Furthermore, the investigated area is suffering
represent a major development in machine learning from water shortage, thus farmers frequently use drainage
algorithms (Ivanciuc, 2007). Also, it could separate the water for supplementary irrigation officially or unofficially.
classes with a decision surface that maximizes the margin Therefore, its vital to have a consistent monitoring strategy
between the classes. That surface is called the optimal to keep track of the changes that is consistently occurring
hyper plane and the data points closest to the hyper plane in this area.
are called support vectors. That hyper plane represents the
optimal separation of linearly separable classes in decision
boundary space (Szuster et al., 2011; Thai et al., 2012 and
Moughal, 2013).SVM is a non-parametric binary classifier
that does not require the assumption of normality and is
812. Glo. Adv. Res. J. Agric. Sci.
vegetation, fish ponds and settlements. The separability of SVM classifiers were 80.28% and 80.64%, respectively.
the training sets, represented by region of interests (ROI), Thus indicating that the classification ability of both
classifiers were the same when used in land use/ land
was evaluated to examine spectral separability between cover classification in this study area. In accordance, the
each two classes. The more separable training data is the Kappa coefficient were almost the same (0.764for MLC
better classification is expected that relies on that data to and 0.768 for SVM), indicating a similar agreement
be. The separability between classes was always about between the classification and ground truth data in case of
1.9(more 1.88) according to both Jeffries-Matusita and MLC and SVM. The output maps from the MLC and SVM
Transformed Divergence measures thus indicating that classification techniques are shown in Map 2 and 3,
each class is properly separated from each other. respectively.
Three classification techniques were adopted in this The ANN classifier attended only an accuracy of 23.33%
study to classify the land use/ land cover classes, namely after various attempts to enhance the accuracy by
MLC,SVM and ANN, and tested using the confusion alternating the neural network parameters. Trying to
matrix. The overall classification accuracy from MLC and
increase the accuracy of the ANN classifier, the data were better results when classifying the settlements class with
processed using Pattern recognition tool of the Neural percentage of 70.95%, while with the SVM this class was
Network application of the Matlab software but still the only 39.59% and was mostly misclassified as sandy bare
results were not satisfactory. The overall accuracy did not soils (49.59%).
exceed 65 % and extremely misclassified two classes in The producer and user accuracy of the MLC and SVM
particular the fish ponds and the sandy soils classes (Map are shown in figure 1 a and b, respectively. Based on these
4).Therefore, the results of the ANN were discarded in this results, the odd classification behavior for the settlements
study and not discussed further. class classified by the SVM clas ifier could be observed. It
Based on the results of the confusion matrix of MLC showed a very low producer accuracy accompanied by a
(Table 1) and SVM (Table 2), it could be concluded that the very high users accuracy. This classifier behavior is not
water and settlements are the most critical classes in MLC desirable. Therefore, contrary to the previous studies
and SVM respectively. Only 59.46% of the water class was (Moughal, 2013 and Wong and Sarker,2014), it is
classified correctly using the MLC, while the rest of the recommended to use the MLC for land use/ land cover
pixels in the class were miss-classified as fish ponds classification in the study area especially when considering
(20.95%) and aquatic vegetation (17.23%). On the other the producer and user accuracy and taking into account
hand, the same class showed better results using the SVM that the settlements class is the most changeable class in
with a percentage of 95.27%. Contrary, the MLC showed the study area that needed to be monitored continually.
a b
Figure 2:Producer and user accuracy after including the NDWI.(a. MLC and b.NDWI)
The SVM classifier showed a high producer accuracy and This work also indicated that even so the ANN was not
low user accuracy and consequently, favoring the appropriate for this study but a considerable improvement
classification of the MLC to be used in land use/ land cover in the accuracy could be attained by using Matlab instead
classification in the study area. of ENVI and therefore it is recommended that further
studies should be applied is this aspect.
CONCLUSION
REFERENCES
In this work the Land Use/ Land Cover classification Abdikan S, Bilgin G, Sanli FB, Uslu E, Ustuner M (2015).
accuracy of remote sensing data using Maximum "Enhancing land use classification with fusing dual-polarized Terra
Likelihood (MLC), Artificial Neural Network (ANN) and SAR-X and multispectral Rapid Eye data", J. Appl. Remote Sens.
9(1):096054 (May 08, 2015); http://dx.doi.org/10.1117/1.JRS.9.096054.
Spectral Angle Mapper (SAM) have been compared.
Ahmad A, Quegan S (2012). Analysis of Maximum Likelihood
Attaining this objective also included the selection of an Classification on Multispectral Data. Applied Mathematical Sciences, 6
appropriate extracting features method which in this work (129), 6425 - 6436.
included using Principal Component Analysis (PCA) and Ashish D (2002). Land-Use Classification of Aerial Images Using Artificial
Neural Networks, Master Thesis, Graduate Faculty of The University of
the use of the Normalized Difference Water Index (NDWI). Georgia, Athens, Georgia.
The resulted data from this study recommended the MLC is Carrao H, Goncalves P, Caetano M (2008). Contribution of multispectral
more effective than the SVM in terms of classification and multitemporal information from MODIS images to land cover
overall accuracy and producer-user accuracy. classification. Remote Sensing of Environment, 112 (3): 986997.
818. Glo. Adv. Res. J. Agric. Sci.
Dell Inc., (2013). Statistics: Methods and Applications. Tulsa, OK: Dell. Netherlands.Qiu F, Jensen JR (2004). Opening the black box of neural
(Electronic Version). http://documents.software.dell.com/statistics/current/textbook. networks for remote sensing image classification , INT. J. Remote
Ivanciuc O (2007). Applications of Support Vector Machines in Chemistry Sensing, 25 (9),17491768
Reviews. Computational Chemistry, vol.23, 291-400. Edited by K. B. Santiesteban IA (2003). Multi-sensor image fusion applied to
Lipkowitz and T. R. Cundari. Wiley-VCH2007, John Wiley & Sons, Inc. morphodynamics of coastal environments in Belize. M.Sc. Thesis,
Japan Association of Remote Sensing (JARS), (1999). Remote Sensing International Institute for Aerospace Survey and Earth Sciences (ITC),
Notes, Chapter 11, Image Processing - Classification. (Electronic Enschede, Netherlands.
version).http://wtlab.iis.u-tokyo.ac.jp/ ~wataru/lecture/rsgis/rsnote/contents.htm Shalaby A, Tateishi R (2007). Remote sensing and GIS for mapping and
Jensen JR (2005). Introductory Digital Image Processing: A Remote monitoring land cover and land-use changes in the Northwestern
Sensing Perspective, Third ed. Prentice Hall, Upper Saddle River, New coastal zone of Egypt. Applied Geography, 27 (1): 2841.
Jersey. Szuster BW, Chen Q, Borger M (2011). A comparison of classification
Karsli F, Gunerogluand A, Dihkan M (2001). Spatio-temporal shoreline techniques to support land cover and land use analysis in tropical
changes along the southern Black Sea coastal zone. J. Appl. Remote coastal zones. Applied Geography 31 (2): 525-532
Sens. 2011, 5, 053545. Thai LH, So T, Hai, Thuy NT (2012). Image Classification using Support
Manandhar R, Odeh OAI, Tiho A (2009). Improving the Accuracy of Land Vector Machine and Artificial Neural Network. I. J. Information
Use and Land Cover Classification of Landsat Data Using Post- Technology and Computer Science, 2012, 5, 32-38
Classification Enhancement. Remote Sensing, 1, 330-344.
Wong SN, Sarker MLR (2014). Land use/ land cover mapping using multi-
Mcfeeters SK (1996). The use of the normalized difference water index
scale texture processing of high resolution data. The 8th International
(NDWI) in the delineation of open water features. Int. J. Remote Sens.,
Symposium of the Digital Earth (ISDE8), IOP Conf., Series: Earth and
17, 14251432.
Environmental Science 18 (2014) 012185.
Miller DM, Kaminsky EJ, Rana S (1995). Neural Networks Classification of
Yuan H, Van Der Wiele CF, Khorram S (2009). An Automated Artificial
Remote Sensing Data. Computers and Geosciences Volume 21 (3):
Neural Network System for Land Use/Land Cover Classification from
377386.
Landsat TM Imager. Remote Sensing, 1, 243-265.
Moughal TA (2013). Hyperspectral image classification using Support
Vector Machine. 6th Vacuum and Surface Sciences Conference of
Asia and Australia (VASSCAA-6). Journal of Physics: Conference
Series 439 (2013) 012042.
Ndehedehe C, Ekpa A, Simeon O, Nse O (2013). Understanding the Neural Network
Technique for Classification of Remote Sensing Data Sets. New York Science
Journal, 6 (8) : 26-33.
Pacifici F, Chini M, Emery WJ (2009). A neural network approach using
multi-scale textural metrics from very high-resolution panchromatic
imagery for urban land-use classification. Remote Sensing of
Environment, 113 (6): 12761292.
Pemberty JFP (2002). Assessment of Morphodynamic Processes and Soil
Pollution as Indicators of Land Degradation. A Case Study in the
Tabernas Area, Southeastern Spain. M.Sc. Thesis, International
Institute for Aerospace Survey and Earth Sciences (ITC), Enschede,