Built a program that searched a greyscale image for Waldos
Was able to detect normal waldos, blurred waldos, waldos with noise, double-sized waldos and double-sized blurred waldos while ignoring objects that were not waldos
Built a binarize function to help the process
Built a neural net using Matlab to recognize the difference between hedgehogs and porcupines
After tweaking parameters, got more than 85% accuracy
Built training and testing sets for the neural net
Built a program to detect the distance between a camera and an object (a simple fiducial)
Binarized the image, using a Gaussian and a median filter to clear salt-and-pepper noise
Segmented the image using a self-built algorithm
Detected circles and differentiated from other noise, and found the centroids of the circles
Calculated distance between centroids and compared to a known distance
Used the ratio to compute the distance between the camera and the fiducial
Got within 5% of accepted values