|
|
Map machine recognition and optimal path planning
Peška, Jaroslav ; Ligocki, Adam (referee) ; Chromý, Adam (advisor)
This masters thesis continues in the work of two previous theses dealing with machine understanding of maps and modelling terrain. The final program also has to be able to interpret position data from dataloggers and integrate it with the loaded map. The goal for the program is to serve during training of Czech orienteering runners. Position measurement and storage is researched first. Also researched are map markers used to define the terrain. Afterwards, past approaches are evaluated, including identification of most severe issues hindering the usage in real world applications. Many improvements are proposed, for example methods to remove noise in the input data, or to improve processing speed. Lastly, a set of possible improvements to the original applications are made, i.e. methods for denoising the input data or for speedup of the image processing. Proposed improvements are then implemented, the most impactful being processing speed and contour segmentation improvements.
|
|
|
Road detection using data from mobile robot camera
Peška, Jaroslav ; Horák, Karel (referee) ; Petyovský, Petr (advisor)
Bachelor thesis deals with the problem of road detection by mobile robots using data obtained from its camera. First, current solutions are researched and considered for use in the proposed algorithm. Afterwards we define limit parameters of the entire solution. An automatic process using human-created reference was then devised to programatically determine the accuracy of individual versions of proposed solutions. First, an initial version of the solution was implemented, which was subsequently optimized and accelerated using GPGPU. Lastly, proposed algorithm is evaluated and possible future changes are outlined.
|
|
|
Map machine recognition and optimal path planning
Peška, Jaroslav ; Ligocki, Adam (referee) ; Chromý, Adam (advisor)
This masters thesis continues in the work of two previous theses dealing with machine understanding of maps and modelling terrain. The final program also has to be able to interpret position data from dataloggers and integrate it with the loaded map. The goal for the program is to serve during training of Czech orienteering runners. Position measurement and storage is researched first. Also researched are map markers used to define the terrain. Afterwards, past approaches are evaluated, including identification of most severe issues hindering the usage in real world applications. Many improvements are proposed, for example methods to remove noise in the input data, or to improve processing speed. Lastly, a set of possible improvements to the original applications are made, i.e. methods for denoising the input data or for speedup of the image processing. Proposed improvements are then implemented, the most impactful being processing speed and contour segmentation improvements.
|
|
|
Road Detection Using Data From Mobile Robot
Camera
Peška, Jaroslav
The paper is focused on developing a road detection algorithm that uses only data from a mobile robot’s camera. Key requirements are low latency and relatively low power requirements. Presented algorithm makes use of machine learning, where the neural network is fed not only image data, but also select additional inputs.
|
|
|
Road detection using data from mobile robot camera
Peška, Jaroslav ; Horák, Karel (referee) ; Petyovský, Petr (advisor)
Bachelor thesis deals with the problem of road detection by mobile robots using data obtained from its camera. First, current solutions are researched and considered for use in the proposed algorithm. Afterwards we define limit parameters of the entire solution. An automatic process using human-created reference was then devised to programatically determine the accuracy of individual versions of proposed solutions. First, an initial version of the solution was implemented, which was subsequently optimized and accelerated using GPGPU. Lastly, proposed algorithm is evaluated and possible future changes are outlined.
|
guest ::
RSS feed.