Finding and marking the contours of objects in an image

Računalni vid predstavlja jedno od najbrže rastućih područja umjetne inteligencije. Temelj računalnog vida predstavlja prepoznavanje kontura objekata sa slike. Cilj ovog rada je predstaviti dva najčešće korištena algoritma za prepoznavanje kontura, a to su Cannyjev algoritam i thresholding metoda. Kroz rad dana je teorijska osnova oba algoritma te je izrađena programska aplikacija za usporedbu navedenih metoda. Aplikacija je izrađena pomoću Python programskog jezika i OpenCV biblioteke. Usporedba metoda vrši se na temelju broja pronađenih kontura i vremenu potrebnom da se program izvrši. Uz to, analizirane su prednosti i nedostatci odabranih metoda.Computer vision is one of the fastest growing areas of artificial intelligence. The basis of computer vision is the recognition of object contours in images. The aim of this paper is to present the two most commonly used contour recognition algorithms, namely the Canny algorithm and the thresholding method. The paper outlines the theoretical basis of both algorithms and the development of a software application for comparing the above methods. The application was created using the Python programming language and the OpenCV library. The methods are compared based on the number of contours found and the time required to execute the program. Additionally, the advantages and disadvantages of the selected methods are analysed