Implementation of lane detection in a video for an autonomous vehicle. I mainly used OpenCV tools to turn the image grayscale, reduce the noise, turn it to gradient image, choose a region of interest, apply bitwise end, and finally hough transformation to identify and display the lines in a video. Full explanation can be found here.
2. MNIST Classification
I created a CNN, trained it on the MNIST dataset, with the aim to classify handwritten digits. The project can be found here. All the subfolders between Course 2 and 7 show the steps and reasoning to reproduce a CNN, developed from a simple perceptron, up to DNN, and then CNN.
3. Road Signs Image Classification
I created a CNN and trained it on road signs images, to be able to classify images that have not been used for training. The project can be found here. I also describe all the image pre-processing needed to reach the final classification.
4. Autonomous Driving
Using all the previous knowledge, I built a CNN model for lines and signs detection that fed a behavioral cloning model, to be able to drive autonomously. More information can be found here.