Denna avhandling beskriver ett projekt för att skapa en prototypapplikation med syftet att förenkla vägmärkesinventering. Istället för att manuellt analysera en tagen bild och jämföra med en databas av vägmärken för inventing så kan man istället starta denna applikation, ladda in bilden och få ut ett svar på vad skylten har för identifieringskod. Idén är att vägmärkesinventerare ska spara in tiden det tar att gå igenom alla bilder tagna under en dag och istället få systemet att automatiskt lägga in vad vägmärket har för identifieringskod. Grundapplikationen är skriven som en WPF-applikation med hjälp av ramverket EmguCV som i sin tur nyttjar .NET ramverket. Den viktiga aspekten i detta projekt är att se ifall detta kan göras med rimlig beräkningskraft, kunna matcha skyltar på en rimlig tid, vilket gjordes möjligt med EmguCVs FLANN-algoritm. Projektet resulterade i en fungerande applikation där användare kan ladda upp cirkulära hastighetsvägmärken där applikationen detekterar och sedan matchar vägmärket mot en databas för att kunna identifiera det. / This dissertation describes a project for creating a prototype application for the purpose of simplifying road sign inventory. Instead of manually analyzing a captured image and comparing it to a database of inventory road signs, you can instead launch this application, load the image and get the identification code of that road sign. The idea is that road sign inventory takers will save the time it takes to review all the pictures taken during a day and instead, the system will automatically generate the identification code of that road sign. The basic application is written as a WPF application using the EmguCV framework, which in turn uses the .NET framework. The important aspect of this project is to see if matching road signs can can be done with reasonable computation and within reasonable time, this was made possible with EmguCVs FLANN-algorithm. The project resulted in a functional application in which users can upload circular velocity road signs and the application detects and identifies the road sign via a database of road signs.

Datorseende är ett område inom datavetenskap som har utvecklats i många år och funktionaliteten är mer tillgängligt nu än någonsin.  Det kan bland annat användas för beröringfri mätning så som att hitta, verifiera och identifiera defekter för objekt. Frågan är om det går att utforma en öppen källkodslösning för datorseende som ger motsvarande prestanda som tillgängliga kommersiella. Med andra ord, kan ett företag som använder ett kommersiellt program med stängd källkod istället använda ett gratis öppet källkodsblibotek och få motsvarande resultat? I denna rapport beskriver vi designen av en prototyp som använder det öppna källkodsbiblioteket för datorseende, OpenCV.  I syfte att utvärdera vår prototyp låter vi den identifiera block i ett torn på en bild i en serie testfall. Vi jämför resultaten från prototypen med de resultat som erhålls med en kommersiell lösning, skapad med programmet 'Vision Builder for Automated Inspection'. Resultat av det som testats visar att OpenCV tycks ha prestanda och funktionalitet som motsvarar den kommersiella lösningen men har begränsningar. Då OpenCV är fokus är på programmatisk utveckling av datorseenden lösningar är resultatet av lösningar som skapas beroende på användarens kompetens inom programmering och programmdesign. Utifrån de tester som genomfördes anser vi att OpenCV kan ersätta ett licensierat kommersiellt program men licensenkostnaderna kan komma att ersättas av andra utvecklingskostnader. / Computer vision is a subject in computer science that have evolved over many years and the functionality is more accessible then ever. Among other things, it can be used for non-contact measurement to locate, verify, and detect defects of objects. The question is if it is possible to create an open source solution for computer vision equivalent to a closed source solution. In other words, can a company using a closed source commercial program instead use a free open source code library and produce equivalent results?  In this report we describe the design of a prototype that uses the open source library for computer vision, OpenCV.  In order to evaluate our prototype, we let it identify block in a tower on image in a series of test cases.  We compare the results from the prototype with the results obtained with a commercial solution, created with the program ''Vision Builder for Automated Inspection''.  Results from the cases tested show that OpenCV seems to have performance and functionality equivalent to the commercial solution but has some limitations.  As OpenCV's focus is on programmatic development of computer vision solutions, the result is dependent on the user's skills in programming and program design.  Based on the tests that we have performed, we believe that OpenCV can replace a licensed commerical program, but the license cost may come to be replaced by other development costs.

This thesis explores the application of Contrastive Language-Image Pre-Training (CLIP), a vision-language model, in an automated video surveillance system for anomaly detection. The ability of CLIP to perform zero-shot learning, coupled with its robustness against minor image alterations due to its lack of reliance on pixel-level image analysis, makes it a suitable candidate for this application. The study investigates the performance of CLIP in tandem with various anomaly detection algorithms within a visual surveillance system. A custom dataset was created for video anomaly detection, encompassing two distinct views and two varying levels of anomaly difficulty. One view offers a more zoomed-in perspective, while the other provides a wider perspective. This was conducted to evaluate the capacity of CLIP to manage objects that occupy either a larger or smaller portion of the entire scene. Several different anomaly detection methods were tested with varying levels of supervision, including unsupervised, one-class classification, and weakly- supervised algorithms, which were compared against each other. To create better separation between the CLIP embeddings, a metric learning model was trained and then used to transform the CLIP embeddings to a new embedding space. The study found that CLIP performs effectively when anomalies take up a larger part of the image, such as in the zoomed-in view where some of the One- Class-Classification (OCC) and weakly supervised methods demonstrated superior performance. When anomalies take up a significantly smaller part of the image in the wider view, CLIP has difficulty distinguishing anomalies from normal scenes even using the transformed CLIP embeddings. For the wider view the results showed on better performance for the OCC and weakly supervised methods.

In this thesis, the development of a stereo camera system for an intelligent tool is presented. The task of the system is to identify and localise objects so that the tool can guide a robot. Different approaches to object detection have been implemented and evaluated and the systems ability to localise objects has been tested. The results show that the system can achieve a localisation accuracy below 5 mm.

Thermal spectrum cameras are gaining interest in many applications due to their long wavelength which allows them to operate under low light and harsh weather conditions. One disadvantage of thermal cameras is their limited visual interpretability for humans, which limits the scope of their applications. In this thesis, we try to address this problem by investigating the possibility of transforming thermal infrared (TIR) images to perceptually realistic visible spectrum (VIS) images by using Convolutional Neural Networks (CNNs). Existing state-of-the-art colorization CNNs fail to provide the desired output as they were trained to map grayscale VIS images to color VIS images. Instead, we utilize an auto-encoder architecture to perform cross-spectral transformation between TIR and VIS images. This architecture was shown to quantitatively perform very well on the problem while producing perceptually realistic images. We show that the quantitative differences are insignificant when training this architecture using different color spaces, while there exist clear qualitative differences depending on the choice of color space. Finally, we found that a CNN trained from day time examples generalizes well on tests from night time.

Training data is the bottleneck for training Convolutional Neural Networks. A larger dataset gives better accuracy though also needs longer training time. It is shown by finetuning neural networks on synthetic rendered images, that the mean average precision increases. This method was applied to two different datasets with five distinctive objects in each. The first dataset consisted of random objects with different geometric shapes. The second dataset contained objects used to assemble IKEA furniture. The neural network with the best performance, trained on 5400 images, achieved a mean average precision of 0.81 on a test which was a sample of a video sequence. Analysis of the impact of the factors dataset size, batch size, and numbers of epochs used in training and different network architectures were done. Using synthetic images to train CNN’s is a promising path to take for object detection where access to large amount of annotated image data is hard to come by.

Thermal spectrum cameras are gaining interest in many applications due to their long wavelength which allows them to operate under low light and harsh weather conditions. One disadvantage of thermal cameras is their limited visual interpretability for humans, which limits the scope of their applications. In this thesis, we try to address this problem by investigating the possibility of transforming thermal infrared (TIR) images to perceptually realistic visible spectrum (VIS) images by using Convolutional Neural Networks (CNNs). Existing state-of-the-art colorization CNNs fail to provide the desired output as they were trained to map grayscale VIS images to color VIS images. Instead, we utilize an auto-encoder architecture to perform cross-spectral transformation between TIR and VIS images. This architecture was shown to quantitatively perform very well on the problem while producing perceptually realistic images. We show that the quantitative differences are insignificant when training this architecture using different color spaces, while there exist clear qualitative differences depending on the choice of color space. Finally, we found that a CNN trained from daytime examples generalizes well on tests from night time.

This master thesis explores the possibility of using generative Adversarial Networks (GANs) to refine labeled synthetic code images to resemble real code images while preserving label information. The GAN used in this thesis consists of a refiner and a discriminator. The discriminator tries to distinguish between real images and refined synthetic images. The refiner tries to fool the discriminator by producing refined synthetic images such that the discriminator classify them as real. By updating these two networks iteratively, the idea is that they will push each other to get better, resulting in refined synthetic images with real image characteristics. The aspiration, if the exploration of GANs turns out successful, is to be able to use refined synthetic images as training data in Semantic Segmentation (SS) tasks and thereby eliminate the laborious task of gathering and labeling real data. Starting off from a foundational GAN-model, different network architectures, hyperparameters and other design choices are explored to find the best performing GAN-model. As is widely acknowledged in the relevant literature, GANs can be difficult to train and the results in this thesis are varying and sometimes ambiguous. Based on the results from this study, the best performing models do however perform better in SS tasks than the unrefined synthetic set they are based on and benchmarked against, with regards to Intersection over Union.

The automotive industry is heading towards more automation. This puts high demands on many systems like Pedestrian Detection Systems. Such systems need to operate in real time with high accuracy and in embedded systems with limited power, memory resources and compute power. This in turn puts high demands on model size and model design. Lately Convolutional Neural Networks (ConvNets) have dominated the field of object detection and therefore it is reasonable to believe that they are suited for pedestrian detection as well. Therefore, this thesis investigates how ConvNets have been used for pedestrian detection and how such solutions can be implemented in embedded systems on FPGAs (Field Programmable Gate Arrays). The conclusions drawn are that ConvNets indeed perform well on pedestrian detection in terms of accuracy but to a cost of large model sizes and heavy computations. This thesis also comes up with a design proposal of a ConvNet for pedestrian detection with the implementation in an embedded system in mind. The proposed network performs well on pedestrian classification and the performance looks promising for detection as well, but further development is required.

No description available.