| Title |
Objekto suradimo scenoje identifikavimo tikslumo tyrimai, tikslumo gerinimas prie tų sąlygų, kurioje jis žemas / |
| Translation of Title |
Evaluation of environment identification methods using a smartphone camera. |
| Authors |
Tekorius, Mantas |
| Full Text |
|
| Pages |
76 |
| Keywords [eng] |
image recognition ; object recognition ; 3d ; face recognition |
| Abstract [eng] |
There are a lot of technologies and solutions which makes our lives easier and simpler. Solutions and software like image recognition are becoming more and more popular and used in our daily routine. Image recognition technologies are in video games, augmented reality, security, robotics and military industry. Most of the systems are already in production but solutions for ordinary people just recently appeared. For their popularity we should thanks smart phones which made huge impact in our lives. This paper concentrates on various image recognition methods, algorithms and approaches and use it in working system. System was designed and using OpenCV library a solution for Android device was created. App is capable to detect humans in camera view and report about it to user with 79% accuracy. While doing a research it was found that HAAR Cascade classifier is most accurate for people face recognition. It was found that scale factor value for best results is 1.05 and minimum neighbors value is 4. In research were used different people face photos and also system were used in real life scenario. |
| Dissertation Institution |
Kauno technologijos universitetas. |
| Type |
Master thesis |
| Language |
Lithuanian |
| Publication date |
2017 |
