AI

CH-3 – Image Analysis – Part-5

Wavelet feature extraction involves transforming images into wavelet coefficients. The number of data elements per image is determined based on the image’s rows and columns, typically brought to a power of two. The total number of values after applying the wavelet is calculated. Classification employs a Dual Statistical Approach, combining parametric and non-parametric methods like MEAN, NORM, Variance, and KS TEST. The Minimum Distance Classifier, utilizing Euclidean, Normalized Euclidean, and Mahalanobis distances, aids in classifying unknown image data by minimizing distances between data and classes in multi-feature space.

AI

CH-3 – Image Analysis – Part-4

Edge feature extraction involves two elements: Edge Map and Edge Direction. The Edge Map provides an image’s edge representation, following the Canny Edge Detector Algorithm. Edge Direction divides the total direction into four parts, calculated using Gaussian Values. Slope feature extraction comprises Slope Magnitude, Slope Direction, and Slope Signs. Slope Magnitude calculates maximum intensity change, Slope Direction determines the direction of this change, and Slope Signs categorize values as positive, negative, or zero, useful for classification. All features are calculated tile-wise and arranged into arrays for analysis, enabling detailed image characterization.

AI

CH-3 – Image Analysis – Part-3

Parametric data follows a specific distribution with consistent distances from the mean, while non-parametric data lacks a specific distribution. Mean frame histogram data, being non-parametric, employs the KS Test. Four features are chosen for extraction: Histogram, Edge, Slope, and Wavelets. Histogram feature extraction involves arranging data in columns and addressing the issue of identical mean values in different distributions by weighting the matrix with a column vector. This process ensures unique results, crucial for decision-making, as mean and variance pairs differ, offering valuable insights.

AI

CH-3 – Image Analysis – Part-2

Descriptive statistics were developed to reduce the list of all the data items to a few simpler numbers. For example, the mean, median, high, low, and standard deviation. The cumulative fraction function displays how the data is distributed, with most data clustered on the left indicating a non-normal distribution. Scaling the x-axis, typically using a log scale for positive data, allows for better visualization. The KS-test computes the maximum vertical deviation between two datasets’ cumulative fraction curves, indicating differences in distribution.