Coursera Project: Applying Optical Flow to Detect Moving Objects: Counting Cars

Project: Applying Optical Flow to Determine the Direction of Traffic

Coursera content: Object Tracking and Motion Detection with Computer Vision

You previously developed code to apply optical flow to a few frames from a video of a highway and created a mask of the fastest-moving objects:

Here, you will use optical flow and this mask to determine how many cars are traveling in each direction and their velocity.

In this problem, you will:

• Copy your code from the previous problem to calculate the optical flow and create a flow magnitude mask.
• Calculate the average for each object in the mask.
• Store the average for every object moving to the left and has an average above 3 in a variable named leftVx. These are your cars driving left.
• Store the average for every object moving to the right and has an average above 3 in a variable named rightVx. These are your cars driving right.
• Store the total number of cars driving to the left as numCarsLeft.
• Store the total number of cars driving to the right as numCarsRight.
• You will be graded based on the following:
• Correctly detecting the number of left-moving vehicles, numCarsLeft
• Correctly detecting the number of right-moving vehicles, numCarsRight
• Correct values for the array of left-moving velocities, leftVx
• Correct values for the array of left-moving velocities, rightVx
• Course link: https://www.coursera.org/learn/intro-computer-vision