Learn Image Processing like Engineers do ! | Online Course

Explore how quantization maps continuous signals to a smaller set of output values, using sampling and an analog-to-digital converter to transform voltages into discrete binary sequences.

Learn how a digital image is represented as a matrix of real numbers, using sampling and quantization to map coordinates and rgb color channels to 0–255 brightness values.

Discover how to display images in Matlab with imshow and image show, adjust contrast using low and high, load images with imread, and compare multiple figures to present enhanced results.

Write images in Matlab using imwrite to save a diagonal image from a matrix, choose formats like jpeg, save to desktop, and preview with imshow.

Learn to write MATLAB functions using the function syntax with inputs and outputs, and comments. Save, run, and test sums with a for loop to support image processing tasks.

Learn how to convert between image classes, such as unsigned integer 8 to a logical black and white image, using im2bw and thresholding before applying filters in MATLAB.

Explore array indexing in digital image processing by building a MATLAB script that draws a letter X using diagonal and flipped diagonal patterns.

Undertake project 1 to create an image of the capital letter A using matrix concepts, with code, hints, and a hands-on exploration of inverse matrices and indexing.

Learn a math-based method to draw graphics with analytical geometry equations, creating a centered white disk using padding and MATLAB in place of per-pixel manipulation.

Design and implement the wowface emoji using analytic equations to render shapes, explore grayscale intensities from 0 to 1 and color schemes, with hints and sample code to guide project.

Explore histogram processing in digital image processing by analyzing gray intensity levels, using 256 bins, MATLAB imhist, and normalization to 0–1 for probability-based analysis.

Explore image enhancement with histogram equalization in MATLAB using the histeq function to map intensities to 256 levels, boosting contrast and dynamic range.

Explore adaptive histogram equalization with adapthisteq, which processes image tiles via linear interpolation and adjustable numtiles to sharpen details, reduce noise, and improve edge clarity in knee x-rays.

Explore neighborhood processing and spatial filtering using masks to apply linear or nonlinear operations to each pixel, and compare correlation with convolution while applying Matlab imfilter to create glow effects.

Demonstrate how spatial filtering uses a one-dimensional mask with zero padding, slides across an input to form full and same correlations, and contrasts correlation with 180-degree rotated convolution.

Learn to blur license plate images with the colfilt filter by converting RGB to grayscale, tuning blur direction and intensity, and finalizing output types for privacy protection.

Explore how frequency filtering enables image enhancement and compression using the 2D discrete Fourier transform. Understand the Fourier spectrum, modulus, phase, inverse transform, and conjugate symmetry for real images.

Code a one-dimensional discrete Fourier transform in Matlab from scratch as an educational exercise, using isvector, numel, and for loops, noting its slow performance compared with fft.

Apply the discrete Fourier transform to a cosine signal, explore time and frequency ranges, and visualize the magnitude spectrum using fftshift, sampling frequency, and frequency in hertz.

Apply logarithmic transformation to reduce the Fourier transform's dynamic range for 8-bit images using c*ln(1+F), then mat2gray restores 0-255 intensities.

Explore computing and visualizing the two-dimensional discrete Fourier transform of an image, using padding and fftshift to center the spectrum and address wraparound, then applying logarithmic transformation and arctan2 phase.

Apply convolution theorem to perform filtering by convolving an image with a mask via the inverse Fourier transform of product of Fourier transforms, and use zero padding to prevent wraparound.

Implement the paddedsize function to compute image padding using AB and CD, with a power-of-two parameter to accelerate the fast Fourier transform.

Follow steps for DFT filtering: pad, compute Fourier transform, generate and apply a filter, shift to avoid wraparound, inverse transform, crop to original size.

Apply Sobel edge detection and frequency filtering in MATLAB, comparing spatial and frequency-domain results, using ifftshift, dftfilt, and standard image padding techniques.

Explore how low pass filters, including Gaussian, ideal, and Butterworth types, use distances from the center and cutoff frequency D0 to attenuate high frequencies in digital image processing.

Deploy high pass filters to sharpen images by attenuating low frequencies while preserving high frequencies above a cutoff frequency D0, contrasting with low pass filtering that blurs.

Apply a high-pass filter to a low-contrast blood cell image using a Gaussian mask and padding, transforming via dftfilt to emphasize high frequencies while removing the dc component.

Explore high frequency emphasis in image processing by applying high-pass filters, offset, and a multiplier to compensate loss of gray tonality, demonstrated in Matlab.

Explore bandpass and bandreject filters in image processing, built from matching low-pass and high-pass filters; learn their frequency-domain regions, transfer functions, and Butterworth and ideal examples.

Explore bandpass and bandreject filters in image processing using Matlab, including ideal, Butterworth, and Gaussian designs. Visualize filter masks with fftshift for centered transfer functions and understand D0 and W.

Hybrid images blend the low spatial frequencies of one image with the high spatial frequencies of another, producing a view-dependent interpretation.

Investigate hybrid images by combining two photos using high pass and low pass frequency filters, illustrated through the Albert Einstein and Marilyn Monroe anecdote.

learn how geometric transformations map input pixel coordinates to output coordinates and form affine transformations using matrix a and vector b, combining translation with scaling, rotation, and reflection.

Explore the scale operator as a geometric transformation to zoom images, and reveal more details through replication or interpolation, including forward zooming and backward image reduction.

Compare random neighborhood sampling and mean-based interpolation for image shrinking, using a 4x4 matrix and 2x2 blocks to produce a half-size result.

Apply the local neighborhood technique to shrink an image by a factor, selecting the upper-left pixel of each neighborhood to create a 215×215 output from a 430×430 input.

Apply image rotation using a rotate image function by a given degree, convert color to grayscale, and double the input image size for center rotation while handling errors with try-catch.

Apply reflection to reconstruct the missing half of an animal face using a vertical-axis, color image approach. Test the provided code to see the symmetry and explore other orientations.