1. Create a facial recognition program that can be run from Ubunutu Linux using PHP's shell_exec to process an FTP folder of images (faces) against
an OpenCV library of images for facial recognition. The program will output the name of the input image and the result into a MySQL database.
The program should allow the user to pass the FTP folder, i.e., "/var/www/html/input_images/" from the linux command line or pull the path to the
images from a text file.
2. Create a program that creates a Haar Training of facial images to facilite building a library of images used to compare against other images for
face detection/facial recognition. Program should process images located in an FTP folder. Program will be called from PHP
How it works
The user uses a Flash program to upload multiple images to an FTP site/folder. These are the images that will be processed for facial recognition
against a library of images or processed for library images or processed as library images to be used for comparison. The user selects from one of
two buttons to "Process Library Images" or "Facial Recognition". If the user presses "Facial Recognition" the uploaded images are processed for
facial recognition and the results are displayed in a grid within the app. The grid has three columns, Timestamp, image name, detection results.
Timestamp is automatically generated by MySQL, image name and detection results are passed from the custom program described above.
Both custom programs should easily processes 100 images or more. The program should demonstrate 15-20 images uploaded of movie stars, politicians,
or celebrities. These uploaded images should be able to be Haar trained by OpenCV and stored in a folder containing the trained images. Secondly,
the second program should process another folder of images that have been uploaded to an FTP folder that will be compared and scored against the
images processed by the first program. The results of the second program are stored to the MySQL database as described above.
Program can be written in C/C++, Python or Bash script. Both programs should execute from the Ubuntu Linux command-line. Docker may not be used
unless scripted from host command-line.
Programmer will be given SSH and FTP access to an AWS EC2 Ubuntu Linux instance. The instance comes preloaded with OpenCV (docker version), Apache2,
Python, MySQL and PHP.
Please feel free to make suggestions or give us your feedback regarding this project.