Jawaharlal Nehru Technological University Hyderabad B.Tech IV DIGITAL IMAGE PROCESSING Previous Question Paper (July/August-2022 - R18)

Image processing tasks are categorized into three levels based on the complexity of the operations performed:

1. Low-level processing: Involves primitive operations such as noise reduction, contrast enhancement, and image sharpening. The input and output are both images, and the process is typically automated.
2. Mid-level processing: Involves tasks like segmentation, description, and object recognition. The input is an image, but the output consists of attributes extracted from the image, such as edges, contours, or object identities.
3. High-level processing: Involves 'image analysis' or 'computer vision', where the system performs cognitive functions similar to human vision, such as interpreting scenes, understanding context, and making decisions based on recognized objects.

The Walsh transform is based on the Walsh functions, which take values of only +1 or -1. The basis functions are defined as a product of Rademacher functions. For a discrete sequence of length N=2^n, the Walsh basis functions g(i, x) are defined as:

where i is the frequency index and x is the spatial index, both ranging from 0 to N-1. The term b_k(i) represents the k-th bit in the binary representation of the index i. The basis functions are orthogonal and symmetric, forming a matrix where each row represents a Walsh function of a specific sequency (the number of zero-crossings).