2. Feature extraction

To classify/segment the different objects in a digital image, features are of much important. Texture feature is one such important feature. Texture is more useful as it is expressed in terms of smoothness, coarseness, fineness, linearization, granularity, and randomness. Analysis of texture requires the identification of features that will differentiate the textures for classification, segmentation, and recognition [17–19, 22, 23, 26, 35–37, 43]. Scale is another important property of texture. The appearance of texture changes when it is viewed at different resolutions. Remotely sensed images are analyzed using gray-level co-occurrence features, features extracted from Gabor filter. There are many methods for extracting features.

#### 2.1 Extraction of features using wavelet packet transform

The main reason for the usage of such wavelet-based multi-resolution analysis [7–10, 12, 27, 29, 30, 39] in remote sensing is that the resolution of the remotely sensed imagery may be different in many cases and it is important to understand how information changes over different scales of imagery.

The work in [1] proposed a system in which statistical and co-occurrence features of the input patterns are first extracted, and those features are used for classification [11–13, 20, 38, 48]. The continuous wavelet transform of a 1-D signal f(x) is defined using Eq. (1):

$$\mathcal{W}(a,b) = \begin{cases} f(\mathbf{x})\Psi\_{a,b}(\mathbf{x})d\mathbf{x} \\ \end{cases} \tag{1}$$

where <sup>Ψ</sup>a,bð Þ¼ <sup>x</sup> <sup>1</sup>ffiffi <sup>a</sup> <sup>p</sup> <sup>Ψ</sup> <sup>x</sup>�<sup>a</sup> b � ). Land Cover/Land Use Mapping Using Soft Computing Techniques with Optimized Features DOI: http://dx.doi.org/10.5772/intechopen.86218

The mother wavelet y has to satisfy the admissibility criterion to ensure that it is a localized zero mean function [39]. Typically, some more constraints are imposed on y to ensure that the transform is non-redundant and complete and constitutes a multi-resolution representation of the original signal. This results in a good realspace transform implementation using quadrature mirror filters. The convolution is performed, and the results with the low-pass filter are called approximation image, and the results with the high-pass filter in specific directions are called detail images. In earlier processes, the image is split into an approximation and detail images. The approximation is then split itself into a second level of approximation and details. For a n-level, the signal decomposition can be represented using Eq. (3):

$$\begin{aligned} \mathbf{A\_{n}} &= \left[\mathbf{H\_{x}} \ast \left[\mathbf{H\_{y}} \ast \mathbf{A\_{n-1}}\right] \downarrow \mathbf{1\_{2,1}}\right] \downarrow \mathbf{1\_{1,2}} \\ \mathbf{D\_{n1}} &= \left[\mathbf{H\_{x}} \ast \left[\mathbf{G\_{y}} \ast \mathbf{A\_{n-1}}\right] \downarrow \mathbf{1\_{2,1}}\right] \downarrow \mathbf{1\_{1,2}} \\ \mathbf{D\_{n2}} &= \left[\mathbf{G\_{x}} \ast \left[\mathbf{H\_{y}} \ast \mathbf{A\_{n-1}}\right] \downarrow \mathbf{1\_{2,1}}\right] \downarrow \mathbf{1\_{1,2}} \\ \mathbf{D\_{n3}} &= \left[\mathbf{G\_{x}} \ast \left[\mathbf{G\_{y}} \ast \mathbf{A\_{n-1}}\right] \downarrow \mathbf{1\_{2,1}}\right] \downarrow \mathbf{1\_{1,2}} \end{aligned} \tag{2}$$

where "\*" denotes the convolution operator, "↓2,1" denotes the downsampling along the rows (columns), A0 = I is the original image, and H and G are low-pass and high-pass filters, respectively. I(x, y) is the original image. An is obtained by low-pass filtering and is the approximation image at scale n. The detail images Dni are obtained by band-pass filtering in a specific direction (i = 1, 2, 3 for vertical, horizontal, and diagonal directions, respectively) and thus contain directional detail information at scale n. The original image, I, is thus represented by a set of subimages at several scales: {An, Dni}.

The wavelet packet decomposition offers a richer signal analysis. Here, the split happens for both detail image and approximation image. This results in a wavelet decomposition tree. The details present in detail images are helpful in analyzing texture and discrimination. To characterize a texture, the features derived from detail images are used. The following section discusses the way in which the features from wavelet transformed image to be used for classification.

The filter choice and its order may vary for each application. Here, two levels of wavelet packet decomposition with different wavelet families are done and shown in Figure 1. There is no need to perform a deeper decomposition because, after the second level, the size of images become too small and no more valuable information is obtained. Sixteen wavelet coefficient matrices containing texture information are produced from the second level of decomposition.

In texture training, the known texture images are decomposed using DWPD. To create feature database, a set of WPSF, such as mean and standard deviation, is

Figure 1. A wavelet packet tree.

So, classifying these will result in the understanding of the variety and type of land cover. Another important advantage with forest land cover is identification of very specific habitats and distribution of both individual species and species assemblies. In the case of urban planning, the year-wise RS images are analyzed to find whether the occupation is growing in the right place. While planning the urban area utilization, the government may plan with the RS image, so that the road construction plan, water pipeline construction plan, and power supply connection plan can be made easy. If in case our urban occupation is happening in the vegetation area, then

RS images are also used in water management system to clearly display sediment pollution and oil spills over water bodies and help to monitor the quality of water resources. They are also used in disaster management. In case of natural disaster, risk-prone areas are detected, and risk management is undertaken. When sudden natural disaster happens, it is difficult for humans to collect data at that moment,

The application area also covers the hazard management. As water-related natural hazards occur due to a number of factors, such as structure, drainage, slope, land use, road network, etc., they must be taken into account when assessing the region's instability and potential hazard risks. It is essential because proper hazard management can help us take timely measures to prevent flooding and

The chapter is organized in the following way. Section 2 explains the feature extraction process, Section 3 explains the feature subset selection, Section 4 explains about feature classification, and Section 5 concludes the chapter.

To classify/segment the different objects in a digital image, features are of much important. Texture feature is one such important feature. Texture is more useful as it is expressed in terms of smoothness, coarseness, fineness, linearization, granularity, and randomness. Analysis of texture requires the identification of features that will differentiate the textures for classification, segmentation, and recognition [17–19, 22, 23, 26, 35–37, 43]. Scale is another important property of texture. The appearance of texture changes when it is viewed at different resolutions. Remotely sensed images are analyzed using gray-level co-occurrence features, features extracted from Gabor filter. There are many methods for extracting features.

The main reason for the usage of such wavelet-based multi-resolution analysis [7–10, 12, 27, 29, 30, 39] in remote sensing is that the resolution of the remotely sensed imagery may be different in many cases and it is important to understand

The work in [1] proposed a system in which statistical and co-occurrence features of the input patterns are first extracted, and those features are used for classification [11–13, 20, 38, 48]. The continuous wavelet transform of a 1-D signal f(x) is defined

f xð ÞΨa,bð Þ x dx (1)

ð

it should be taken care of and constructions are to be made in other areas.

and so using RS technology, we can handle the situation.

2.1 Extraction of features using wavelet packet transform

how information changes over different scales of imagery.

<sup>a</sup> <sup>p</sup> <sup>Ψ</sup> <sup>x</sup>�<sup>a</sup> b � ).

W að Þ¼ ; b

following landslides.

Land Use Change and Sustainability

2. Feature extraction

using Eq. (1):

60

where <sup>Ψ</sup>a,bð Þ¼ <sup>x</sup> <sup>1</sup>ffiffi

calculated to form the original image, and a set of wavelet packet co-occurrence features and spectral feature NDVI is calculated using Eqs. 4–11 and Eq. (12), respectively. These features are saved for further use in texture classification.

$$Mean \quad \overline{\mathfrak{x}} = \frac{1}{N^2} \sum\_{i=1}^{N} \sum\_{j=1}^{N} \mathfrak{x}\_{i,j} \tag{3}$$

$$Variance \quad V = \frac{1}{N^2} \sum\_{i=0}^{N} \sum\_{j=1}^{N} \left( \mathbf{x}\_{i,j} - \overline{\mathbf{x}} \right)^2 \tag{4}$$

$$Entropy = -\sum\_{i=1}^{N} \sum\_{j=1}^{N} \left( \mathbf{C}(i,j) \right) \log \left( \mathbf{C}(i,j) \right). \tag{5}$$

$$\text{Contrast} = \sum\_{i,j=0}^{N} \left( i - j \right)^{2} \mathbf{C}(i, j) \tag{6}$$

$$Energy = \sum\_{i=1}^{N} \sum\_{i=1}^{N} \mathcal{C}(i, j)^2 \tag{7}$$

Figure 2.

Figure 3.

Figure 4.

63

Classified output images using (a) DB2, (b) symlet 2, (c) Coiflet 2, (d) Bi-or 2.2, and (e) DB2 without NDVI.

Land Cover/Land Use Mapping Using Soft Computing Techniques with Optimized Features

Architecture of convolutional neural network. [source: https://images.app.goo.gl/YcBQH2Y4ZXPyMhVr8].

Madurai city (size 400 400).

DOI: http://dx.doi.org/10.5772/intechopen.86218

$$\text{Local homogeneity} = \sum\_{i,j=0}^{n} \mathbf{1} / \left(\mathbf{1} + (i-j)^2\right) \mathbf{C}(i,j) \tag{8}$$

$$\text{Cluster } share = \sum\_{i,j=0}^{n} \left( i - M\_{\text{x}} + j - M\_{\text{y}} \right)^{3} \text{C}(i,j) \tag{9}$$

$$\text{Cluster prominence} = \sum\_{i,j=0}^{n} \left( i - M\_{\text{x}} + j - M\_{\text{y}} \right)^{4} \text{C}(i,j) \tag{10}$$

$$\text{where } M\_{\mathbf{x}} = \sum\_{i,j=0}^{n} i\mathbf{C}(i,j) \text{ and } M\_{\mathbf{y}} = \sum\_{i,j=0}^{n} j\mathbf{C}(i,j).$$

$$\text{Correlation} = \frac{\sum\_{i=1}^{N} \sum\_{j=1}^{N} [j\mathbf{C}(i,j)] - \mu\_{\mathbf{x}}\mu\_{\mathbf{y}}}{\sigma\_{\mathbf{x}}\sigma\_{\mathbf{y}}}. \tag{11}$$

$$\begin{aligned} \text{where } \boldsymbol{\mu\_{x}} &= \sum\_{i}^{N} i \sum\_{j}^{N} \mathbf{C}(i,j), \boldsymbol{\mu\_{y}} = \sum\_{j}^{N} j \sum\_{i}^{N} \mathbf{C}(i,j), \sigma\_{x}^{2} = \sum\_{i}^{N} (\boldsymbol{a} - \boldsymbol{\mu\_{x}})^{2} \sum\_{j}^{N} \mathbf{C}(i,j),\\ \sigma\_{y}^{2} &= \sum\_{j}^{N} \left(\boldsymbol{b} - \boldsymbol{\mu\_{y}}\right)^{2} \sum\_{i}^{N} \mathbf{C}(i,j). \end{aligned}$$

$$\text{NDVI} = (\text{near IR band} - \text{red band}) / (\text{near IR band} + \text{red band}) \tag{12}$$

The input Madurai LISS IV image is shown in Figure 2. The procedure for classification is explained in the later content, but the results are presented here for better understanding. The classification of LISS IV Madurai image is done with wavelet filters such as Daubechies (DB2), symlet (Sym2), Coiflet (Coif2), and biorthogonal (Bi-or2.2) and is shown in Figure 3(a)–(e).

#### 2.2 Extraction of deep features

Deep feature learning plays an important role in image classification. In order to extract different features automatically, the convolution neural network (CNN) is

Land Cover/Land Use Mapping Using Soft Computing Techniques with Optimized Features DOI: http://dx.doi.org/10.5772/intechopen.86218

Figure 2. Madurai city (size 400 400).

calculated to form the original image, and a set of wavelet packet co-occurrence features and spectral feature NDVI is calculated using Eqs. 4–11 and Eq. (12), respectively. These features are saved for further use in texture classification.

> N2 X N

N2 X N

X N

j¼1

N

i,j¼<sup>0</sup>

N

i¼1

i,j¼<sup>0</sup>

i¼1

i¼0

X N

xi,j (3)

xi,j � <sup>x</sup> � �<sup>2</sup> (4)

C ið Þ ; j (6)

C ið Þ ; <sup>j</sup> <sup>2</sup> (7)

C ið Þ ; j (8)

C ið Þ ; j (9)

C ið Þ ; j (10)

: (11)

C ið Þ ; j ,

ð Þ C ið Þ ; j log ð Þ C ið Þ ; j : (5)

j¼1

X N

j¼1

ð Þ <sup>i</sup> � <sup>j</sup> <sup>2</sup>

X N

i¼1

<sup>1</sup><sup>=</sup> <sup>1</sup> <sup>þ</sup> ð Þ <sup>i</sup> � <sup>j</sup> <sup>2</sup> � �

i � Mx þ j � My � �<sup>4</sup>

<sup>j</sup>¼<sup>1</sup> ½ �� ijC ið Þ ; <sup>j</sup> <sup>μ</sup>xμ<sup>y</sup>

<sup>2</sup> <sup>¼</sup> <sup>P</sup> N i

<sup>a</sup> � <sup>μ</sup><sup>x</sup> ð Þ<sup>2</sup> <sup>P</sup>

N j

i � Mx þ j � My � �<sup>3</sup>

jC ið Þ ; j .

σxσ<sup>y</sup>

C ið Þ ; j , σ<sup>x</sup>

NDVI ¼ ð Þ near IR band � red band =ð Þ near IR band þ red band (12)

Deep feature learning plays an important role in image classification. In order to extract different features automatically, the convolution neural network (CNN) is

The input Madurai LISS IV image is shown in Figure 2. The procedure for classification is explained in the later content, but the results are presented here for better understanding. The classification of LISS IV Madurai image is done with wavelet filters such as Daubechies (DB2), symlet (Sym2), Coiflet (Coif2), and bi-

Mean <sup>x</sup> <sup>¼</sup> <sup>1</sup>

N

i¼1

Energy <sup>¼</sup> <sup>X</sup>

i,j¼<sup>0</sup>

i,j¼<sup>0</sup>

i,j¼<sup>0</sup>

P<sup>N</sup> i¼1 P<sup>N</sup>

N j j P N i

Contrast <sup>¼</sup> <sup>X</sup>

Variance V <sup>¼</sup> <sup>1</sup>

Entropy ¼ �<sup>X</sup>

Local homogeneity <sup>¼</sup> <sup>X</sup><sup>n</sup>

Cluster shade <sup>¼</sup> <sup>X</sup><sup>n</sup>

Cluster prominence <sup>¼</sup> <sup>X</sup><sup>n</sup>

Correlation ¼

C ið Þ ; <sup>j</sup> , <sup>μ</sup><sup>y</sup> <sup>¼</sup> <sup>P</sup>

iC ið Þ ; <sup>j</sup> and My <sup>¼</sup> <sup>P</sup><sup>n</sup>

where Mx <sup>¼</sup> <sup>P</sup><sup>n</sup>

where <sup>μ</sup><sup>x</sup> <sup>¼</sup> <sup>P</sup>

b � μ<sup>y</sup> � �<sup>2</sup>

σy <sup>2</sup> <sup>¼</sup> <sup>P</sup> N j

62

i,j¼<sup>0</sup>

Land Use Change and Sustainability

N i i P N j

2.2 Extraction of deep features

P N i

C ið Þ ; j .

orthogonal (Bi-or2.2) and is shown in Figure 3(a)–(e).

Figure 3.

Classified output images using (a) DB2, (b) symlet 2, (c) Coiflet 2, (d) Bi-or 2.2, and (e) DB2 without NDVI.

Figure 4.

Architecture of convolutional neural network. [source: https://images.app.goo.gl/YcBQH2Y4ZXPyMhVr8].

utilized [2]. The architecture of CNN is shown in Figure 4. In convolution layer, the features are extracted using different filters to input image. The ReLU layer handles the output from convolutional layer by figuring out the negative pixel value into zero, retaining the dimensionality of the matrix unchanged. Pooling helps in retaining the most important information while reducing the size of feature map. Each training sample is applied with the same processes and thus resulting in different feature sets.
