The thesis implements a vision-based leader-follower tracking algorithm on a ground robot system. One camera is the only sensor installed the leader-follower system and is mounted on the follower. One sphere is the only feature installed on the leader. The camera identifies the sphere in the openCV Library and calculates the relative position between the follower and leader using the area and position of the sphere in the camera frame. A P controller for the follower and a P controller for the camera heading are built. The vision-based leader-follower tracking algorithm is verified according to the simulation and implementation.

This master's thesis deals with questions about contactless camera detection of presence and right direction of assembled parts in the industry process. The main aim of this work is design and realization of graphical user interface and algorithm for automated camera measurement system in the industrial process.

This work is devoted to the relatively new field of computer – computer vision. It focuses on the recognition of people, positioning and colour detection of clothing placed on person. The aim is to build an algorithm that would be able to locate the person in the picture and would make colours tests of clothing and helmets. For image processing were used OpenCV library functions and from algorithms was compiled program solving this problem. The output of the program is the answer, what colour is person at stated locations wearing, and if clothing and helmet are the same colour, the person is evaluated as properly dressed. The resulting program is then disassembled and parts of the code are in detail described in this work. There is explained how to use correctly each OpenCV function used in program.

Diabetic retinopathy and age-related macular degeneration are two of the most common retinal diseases in these days, which can lead to partial or full loss of sight. Due to it, it is necessary to create new approaches enabling to detect these diseases and inform the patient about his condition in advance. The main objective of this work is to design and to implement an algorithm for retinal diseases classification based on images of the patient's retina of previously mentioned diseases. In the first part of this work, there is described in detail each stage of each disease and its the most frequent symptoms. In this thesis, there is also a chapter about fundus camera, which is a tool for image creation of human eye retina. In the second part of this thesis, there is proposed an approach for classification of diabetic retinopathy and age-related macular degeneration. There is also a chapter about algorithmic methods which can be used for image processing and object detection in image. The last part of this thesis contains the test results and their evaluation. Assessment of success of proposed and implemented methods is also part of this chapter.

The thesis deals with OpenCV library and its implementation into scripting language Lua. The first part of the thesis concentrates on description of the course Computer vision MPOV and description of mathematical basics needed for further understandings. The second part describes OpenCV library and its potential usage in the MPOV. Next chapter examines the programming scripting language Lua. The description of the implementation of binding the OpenCV library to Lua language along with its overall functionality is included in the practical part of the thesis. The use of LuaCV is more comfortable thanks to Open Source projects for cross-platform compilation and distribution. Part of the thesis is also generator of Latex documentation for LuaCV binding. The last chapter deals with testing LuaCV in course MPOV and analysis of criticism from students.

This thesis deals with content based automatic photo categorization. The aim of the work is to experiment with advanced techniques of image represenatation and to create a classifier which is able to process large image dataset with sufficient accuracy and computation speed. A traditional solution based on using visual codebooks is enhanced by computing color features, soft assignment of visual words to extracted feature vectors, usage of image segmentation in process of visual codebook creation and dividing picture into cells. These cells are processed separately. Linear SVM classifier with explicit data embeding is used for its efficiency. Finally, results of experiments with above mentioned techniques of the image categorization are discussed.