Image Processing | OpenCV

Read and write images with OpenCV, understand image composition with illuminance and reflectance, and choose lossy or lossless compression while displaying via Matplotlib in Colab.

Convert color images to grayscale using the luminosity method with weights 0.2 r, 0.5 g, 0.1 b, reducing computation and memory, and demonstrate OpenCV BGR to grayscale and its benefits.

Learn to blend and subtract images using cv2.addWeighted, creating linear image blends for detail across light levels, noise reduction, and multi-exposure effects, with grayscale displays and histogram visualization.

Explore how bitwise and, or, not, and xor operate on pixel values and masks to mask regions, subtract backgrounds, and blend or watermark images.

rotate an image by 90 degrees clockwise, flip it vertically or horizontally, and boost brightness by converting to grayscale with OpenCV and displaying the original and the result.

Learn how image resizing changes width and height to balance storage or detail, and compare linear interpolation with nearest neighbor using grayscale conversion and cv2.resize.

Learn how image blurring, or image smoothing, reduces noise and detail by convolving with low-pass kernels, using box, Gaussian, median, and bilateral blur in cv2.

Apply image enhancement techniques in both spatial and frequency domains, including gamma correction and histogram equalization, with CLAHE to boost local contrast and edge definition.

Explore noising and denoising models for image and signal processing, including Gaussian and speckle noise, and compare filters such as Gaussian, mean, and median, plus denoising autoencoders and wavelet transforms.

Apply non-linear morphological operations using a structuring element to shape image features, including dilation, erosion, opening, and closing. Learn how these steps process binary and grayscale images and extract structures.

Explore dilation and erosion in OpenCV image processing by applying a 10x10 kernel to grayscale images, comparing eroded, dilated, and original results side by side.

Explore opening and closing, two fundamental morphological operations derived from erosion and dilation, to remove noise, fill holes, and connect elements. Learn how structuring elements and OpenCV implement these techniques.

Explore boundary extraction and region filling using morphological operations in image processing with OpenCV, including erosion, subtraction, inversion, dilation, and contour finding.

Discover contour detection and convex hull computation with OpenCV, from grayscale and binary prep to noise reduction, contour drawing, and hull analysis for shape and object recognition.

Explore image segmentation techniques—thresholding, region-based, edge detection, clustering, watershed, semantic and instance segmentation. Compare Sobel and Canny, gaussian smoothing, gradient, non-maximum suppression, thresholds and edge linking.

Explore image descriptors, including color histograms, texture descriptors such as Haralick features and LBP, shape descriptors, local features like SIFT and SURF, and deep learning descriptors learned by CNNs.

Explore texture descriptors with local binary pattern on grayscale images and extract wavelet-based features through a 2d transform, yielding approximation and detail coefficients for multi-scale texture analysis.

Learn to represent color content with a three dimensional rgb color histogram, measure dominant colors and distribution, adjust with 256-bin histograms using OpenCV, masks, and channel-specific calculations.

Develop and evaluate a deep learning cnn for four-class brain tumor classification, from GitHub data loading and grayscale preprocessing to keras training and comprehensive metrics.