Computer Vision News - April 2018

1. Opening an AWS.amazon account First, go to Amazon Web Services at aws.amazon.com and open an account -- you can open a free account for a trial period. Next, under the Compute category, select Lambda. Once the Lambda window opens, select Create function. Selecting the option Author from scratch opens some fields. We’ll need to Name the function, we chose to call it 1lambdaCVN. For the field Runtime environment we’ll select Python 3.6. Finally, to create the Lambda function press the Create function button at the bottom of the page. Note that “Function” is the name AWS gave each instance of their Lambda service, it does not refer to a function in the computing sense. 2. Setting up a Python work environment that includes OpenCV First -- an important warning -- all of the procedures detailed in this section must be executed only in a Linux environment. This is because AWS runs Linux servers. In our case, we want to use libraries such as Numpy and OpenCV that are not a part of the basic services libraries installed on the Lambda servers. Now, first you need to install Python with the relevant code packages -- we recommend using the Anaconda distribution and the well-known python IDE Pycharm, both of which can be easily downloaded and installed from the given links. After downloading and installing the files, run Pycharm. We’ll use Pycharm to create the code and work environment that we will upload to our Lambda cloud service, and to rename the new work environment: To set up an environment with the appropriate software packages go to File → Setting → Project interpreter and on the button that looks like a gear select Add local. From there, we’ll select the Numpy and OpenCV libraries to install. Here’s what the OpenCV installation will look like: Computer Vision on Amazon Cloud 13 Tool Computer Vision News