supervised classification methods

Maximum likelihood

The maximum likelihood classifier is a prominent remote sensing supervised classification approach in which a pixel with the highest likelihood is classified into the appropriate class.

Maximum likelihood classification determines the probability that a given pixel belongs to a specific class based on the statistics for each class in each band being normally distributed. All pixels are categorized unless a probability threshold is set.
The class with the highest probability is assigned to each pixel (that is, the maximum likelihood). The pixel stays unclassified if the highest likelihood is less than a threshold you specify.

The advantages

From the perspective of probability theory, the maximum likelihood method has an advantage, but there are a few things to keep in mind.
1) Enough ground truth data should be sampled to allow for estimation of the population’s mean vector and variance-covariance matrix.
2) When there is a lot of correlation between two bands or the ground truth data is very homogeneous, the inverse matrix of the variance-covariance matrix becomes unstable. In such circumstances, a principal component analysis should be used to limit the number of bands.
3) The greatest likelihood method cannot be used when the population distribution does not follow a normal distribution.

Support vector machine (SVM)

It is an excellent algorithm for categorization. It’s a supervised learning algorithm that’s primarily used to categorize data into several groups. A set of label data is used to train SVM. SVM has the advantage of being able to solve both classification and regression issues. To divide or classify two classes, SVM creates a decision boundary, which is a hyperplane between them. SVM is also utilized in picture classification and object detection.

Support Vector Machines (SVM) are classified as a classification method, however they can be used to solve both classification and regression problems.

It can handle both continuous and categorical variables with ease. To differentiate various classes, SVM creates a hyperplane in multidimensional space.

SVM iteratively generates the best hyperplane, which is then utilized to minimize an error. The goal of SVM is to find a maximum marginal hyperplane (MMH) that splits a dataset into classes as evenly as possible.

It is a binary classification problem, but Support Vector Machine can also be used for multiclass classification problems

Principal Component Analysis

It is feature based image classification

When compared to other techniques, Principal Component Analysis (PCA) improves the accuracy of feature-based image identification and classification.

PCA is a feature-based classification method that is commonly used in image recognition. PCA is based on an image’s main features, which are used to discretely represent the image.