This repository contains three image processing projects utilizing different techniques and algorithms. Below are the details of each project.
In this project, a pre-trained neural network is used to perform style transfer between two images. Style transfer allows combining the content of one image with the style of another, thus creating a new stylized image.
- Model Used: Magenta Arbitrary Image Stylization v1-256
- Framework: TensorFlow Hub
In this project, mathematical morphology techniques are applied to count the number of holes present in a printed circuit board (PCB). Accurate identification of holes is crucial for the analysis and manufacturing of PCBs.
- Techniques Used: Morphological operations such as dilation and erosion.
This project involves segmenting images using the watershed algorithm. Watershed segmentation is particularly useful in situations where objects in an image overlap or are adjacent to each other.
- Technique Used: Watershed transform.
An algorithm was developed to binarize grayscale images. Binarization is a process that converts a grayscale image into a binary image, where the pixels are represented by only two possible values: black and white.
- Objective: Highlight specific features of an image to facilitate subsequent analysis.
Tools to calculate and visualize the cumulative distribution function (CDF) and histograms of images. These techniques are essential for statistical analysis of images and for applying histogram equalization.
Implementation of gamma correction to adjust the brightness of images. Gamma correction is a nonlinear operation used to encode and decode luminance or tristimulus values in images.
Various techniques were applied to augment the size of images while preserving quality. Augmentation techniques are crucial for improving image resolution and for preparing data for machine learning models.
- Techniques Used: Bilinear interpolation, bicubic interpolation, among others.
- Clone the repository:
git clone https://github.com/your-username/your-repository.git