April 20th, 2015.
Today I tested the code for camera, it worked well in normal condition but also has some problems. Firstly, when the camera was moving too fast, it lost the frame and then re-extracted new feature point, as a result it messed up with the calculation and ultimate outcome. I had some ideas about fixing it but there was no efficient solution satisfying the requirements still. Secondly, this is not exactly a problem but it was quite strange. The absolute result in distance is really small regarding to a total distance. Although I had learnd about the accuracy of this method in paper, I did not expect such high accuracy in my self-implemented system because the authors used much more efficient extracting method than I did. I assume that I need to carry out more experiment to conclude the result. In the end, I believe I have done 70% work related to the project, 30% left is about mechanical and improving perfomance of system in the real world.
I have just send my CV to Robert Bosch VN for applying internship program. Honestly, I would like to work in FrieslandCampina as factory engineer but sadly they don’t recruit students who haven’t gradutated yet, therefore I have to pin all my hope on getting offer from Bosch in the midst of hopelessly waiting for Renesas, which I did the test poorly, and Nike which might require some nature that do not belong to a typical engineer. I hope I can pull things off because in the end of the day, what matters is whether I get the intern position by my own.