|
|
Image data segmentation using deep neural networks
Hrdý, Martin ; Myška, Vojtěch (referee) ; Kiac, Martin (advisor)
The main aim of this master’s thesis is to get acquainted with the theory of the current segmentation methods, that use deep learning. Segmentation neural network that will be capable of segmenting individual instances of the objects will be proposed and created based on theoretical knowledge. The main focus of the segmentation neural network will be segmentation of electronic components from printed circuit boards.
|
|
|
Deep Neural Networks for Defect Detection
Juřica, Tomáš ; Herout, Adam (referee) ; Hradiš, Michal (advisor)
The goal of this work is to bring automatic defect detection to the manufacturing process of plastic cards. A card is considered defective when it is contaminated with a dust particle or a hair. The main challenges I am facing to accomplish this task are a very few training data samples (214 images), small area of target defects in context of an entire card (average defect area is 0.0068 \% of the card) and also very complex background the detection task is performed on. In order to accomplish the task, I decided to use Mask R-CNN detection algorithm combined with augmentation techniques such as synthetic dataset generation. I trained the model on the synthetic dataset consisting of 20 000 images. This way I was able to create a model performing 0.83 AP at 0.1 IoU on the original data test set.
|
|
|
Detection of Graffiti Tags in Image
Fischer, Martin ; Kodym, Oldřich (referee) ; Špaňhel, Jakub (advisor)
The aim of this work is to compare different approaches of computer vision with the intention of automatic detection of graffiti tags in the image. The solution was based on models based on neural networks. Both the proven detection models and the experimental models were tested here. The most accurate one (Faster R-CNN) achieved an accuracy of 83% mAP, indicating the suitability of these models to the tag detection problem.
|
|
|
Vehicle Detection in Image and Video
Rozprým, Dalimil ; Juránek, Roman (referee) ; Špaňhel, Jakub (advisor)
The goal of this thesis is comparison of available multiclass detectors abilities to detect road vehicles on purposely created dataset. As multiclass detectors are chosen neural networks for detection and classification of objects in image. Detectors described in this text and used for experimentation are Mask R-CNN, YOLOv4 and YOLACT++. This selection encompasses multiple different architectures and approaches to object detection. Created dataset used for learning and testing is thoroughly described in this text. Detection capability of detectors is tested on images from casual traffic and separately on partially covered objects. The outcome of this thesis is reusable and expandable dataset, measured performance values and their deeper exploration in this text.
|
|
| |
|
|
Occupancy Estimation of a Parking Lot from Images
Aghayev, Raul ; Zemčík, Pavel (referee) ; Herout, Adam (advisor)
The aim of a diploma is to create an application that will work detect vehicles on a video from parking areas and determine the occupancy of parking area, by detecting the cars, saving data about them and count the number of busy slots. This kind of application can substitute sensors in a future whereas the cost of it is much cheaper and it detects the new coming cars in a real-time, in addition with some statistics such as the most popular parking slots or total amount of cars on a parking today
|
|
|
Parking Spot Recognition by Artificial Intelligence
Sicha, Marek ; Koudelka, Vlastimil (referee) ; Kadlec, Petr (advisor)
This thesis deals with the recognition of parking spots in images using artificial intelligence. The goal of the work was to study neural networks and select a suitable network to solve the problem. Python programming language was chosen for implementation and Mask R-CNN convolutional network was selected as a suitable neural network. To train the neural network, a custom dataset was created which contains images captured from street cameras. The trained network was then implemented in a program that easily provides information about available parking spaces in a particular area. The program analyzes images from cameras in parking lots and on streets, identifies the number of available parking spaces, and displays this information on a map.
|
|
|
Segmentation of images from a thermal camera using selected convolutional neural networks
BENEDA, Lukáš
This thesis deals with the issue of instance segmentation of cattle's hoof in thermographic images. The aim of this thesis was to test several solutions and evaluate them. The basis of this thesis is review of existing solutions and state-of-the-art in this field of study, dataset preparation, selection of neural network models and evaluation of the results of individual models. The thesis describes the progress of the work and in the conclusion the individual results are compared and the best solution is evaluated. As results of this thesis are the created dataset of thermographic images of cattle's hoofs and 3 tested and evaluated models of neural networks, from which 2 of them are useable as solution for this issue.
|
|
|
Vehicle Detection in Image and Video
Rozprým, Dalimil ; Juránek, Roman (referee) ; Špaňhel, Jakub (advisor)
The goal of this thesis is comparison of available multiclass detectors abilities to detect road vehicles on purposely created dataset. As multiclass detectors are chosen neural networks for detection and classification of objects in image. Detectors described in this text and used for experimentation are Mask R-CNN, YOLOv4 and YOLACT++. This selection encompasses multiple different architectures and approaches to object detection. Created dataset used for learning and testing is thoroughly described in this text. Detection capability of detectors is tested on images from casual traffic and separately on partially covered objects. The outcome of this thesis is reusable and expandable dataset, measured performance values and their deeper exploration in this text.
|
|
|
Occupancy Estimation of a Parking Lot from Images
Aghayev, Raul ; Zemčík, Pavel (referee) ; Herout, Adam (advisor)
The aim of a diploma is to create an application that will work detect vehicles on a video from parking areas and determine the occupancy of parking area, by detecting the cars, saving data about them and count the number of busy slots. This kind of application can substitute sensors in a future whereas the cost of it is much cheaper and it detects the new coming cars in a real-time, in addition with some statistics such as the most popular parking slots or total amount of cars on a parking today
|
guest ::
