Hitting this point in the internship has shaped me a great deal. I have come to learn so much about Opencv and how to apply it to images and even videos. However, I have hit a wall when it comes to counting people through images. The whole purpose of this internship is to be able to count people on the trails. If I cannot apply the counting algorithms to pictures, there is no use in working on this project at all. Sometimes I sit in front of my laptop wondering how long I will go without making any progress. More time researching is necessary. The new images that were taken are now the new experimental images to apply background subtraction and motion capture. Another standstill happened today. Advice from Matt could have been helpful. I am tired of trying these methods on images. Maybe it is time to switch to video capture. So far, these algorithms have worked on video files but struggled on images.
I thought the hard part was over after the algorithm for background subtraction was developed. However, it looks like the hard part has just begun. Once we can figure out the camera problem, getting static backgrounds from the trails, everything else will be a breeze. A presentation still needs to be made for the other students that are doing undergrad research. Presenting is my forte so I have no problem overcoming this challenge. If anything, this will give me an opportunity to improve on my presentation skills.
I thought the hard part was over after the algorithm for background subtraction was developed. However, it looks like the hard part has just begun. Once we can figure out the camera problem, getting static backgrounds from the trails, everything else will be a breeze. A presentation still needs to be made for the other students that are doing undergrad research. Presenting is my forte so I have no problem overcoming this challenge. If anything, this will give me an opportunity to improve on my presentation skills.
Comments
Post a Comment