DJ SYNAPSE-ML-Developer Tasks
TASK 1 - Image Processing
Using python and the openCV library, draw an outline along the edges of the largest rectangle in the image below. The code should determine the largest rectangle. (Hint: take a look at cv2.findContours, cv2.contourArea, cv2.drawContours)
TASK 2 - Synopsis
COMPUTER VISION
Given a dataset of human face images taken from a dashboard of a car, detect if the driver is drowsy or not. Note that a driver can be classified as drowsy if their eyes are closed more often than open. Describe the preprocessing steps required on the image (if any). Describe techniques you would use to detect the eyes in the entire image. Describe techniques you would use to detect drowsiness in the eyes. End the synopsis with a conclusion which should contain information about why you selected a particular algorithm over other existing solutions.
- Describe the preprocessing steps required on the image (if any).
- Describe techniques you would use to detect the eyes in the entire image.
- Describe techniques you would use to detect drowsiness in the eyes.
- End the synopsis with a conclusion which should contain information about why you selected a particular algorithm over other existing solutions.
Note : All modelling techniques must be described in detail. Mathematical explanations would fetch you brownie points! (code for the same is optional).
Consider this as a supervised learning problem where you have been given images of open and closed eyes for training.
Dataset contains images like the ones provided below :
