Detekcija objekata na fotografijama

U ovome radu opisana je detekcija objekata. Rad ima pet poglavlja. U početku je dana definicija detekcije objekata, uz definicije srodnih zadataka računalnog vida. Potom je opisana primjena detekcije objekata u suvremenom društvu. Nakon toga, u poglavlju tri, opisane su vrste detektora, podjela na jednofazne i dvofazne i njihova temeljna načela, temeljni algoritmi iz dviju faza razvoja: prije i poslije 2014. godine, kada se pojavilo duboko učenje. Na kraju ovog poglavlja uspoređuju se različiti algoritmi po metrikama za brzinu i preciznost. Poglavlje četiri daje opis izrade rada i korištene alate, uz prikaz krajnjeg rezultata detekcije objekta na slici. U zadnjem poglavlju se donosi zaključak rada.This theses describes object detection. It has five chapters. Firstly, the definition of object detection is given, along with the definitions of other related subtasks of computer vision. Next, applications of object detection in modern society are given. After that, in chapter three, types of detectors are described, division of them into one-stage and two-stage detectors and their main principles, main algorithms from two periods of development: before 2014. and after that, when deep learning has arrived. At the end of this chapter, a comparison of algorithms by metrics of precision and speed is given. Chapter four describes how the practical part of this work is made, which tools are used and at the end is presented the result of detection on a single image. In the end, the conclusion is made