<p>An image that has been subject to the out-of-focus phenomenon has reducedsharpness, contrast and level of detail depending on the amount of defocus. Torestore out-of-focused images is a complex task due to the information loss thatoccurs. However there exist many restoration algorithms that attempt to revertthis defocus by estimating a noise model and utilizing the point spread function.The purpose of this thesis, proposed by FLIR Systems, was to find a robustalgorithm that can restore focus and from the customer’s perspective be userfriendly. The thesis includes three implemented algorithms that have been com-pared to MATLABs built-in. Three image series were used to evaluate the limitsand performance of each algorithm, based on deblurring quality, implementationcomplexity, computation time and usability.Results show that the Alternating Direction Method for total variation de-convolution proposed by Tao et al. [29] together with its the modified discretecosines transform version restores the defocused images with the highest qual-ity. These two algorithms include features such as, fast computational time, fewparameters to tune and a powerful noise reduction.</p>

In the master thesis about object detection(OD) using deep convolutional neural network(DCNN), the area of OD is being tested when being applied to infrared images(IR). In this thesis the, goal is to use both long wave infrared(LWIR) images and short wave infrared(SWIR) images taken from an airplane in order to train a DCNN to detect runways, Precision Approach Path Indicator(PAPI) lights, and approaching lights. The purpose for detecting these objects in IR images is because IR light transmits better than visible light under certain weather conditions, for example, fog. This system could then help the pilot detect the runway in bad weather. The RetinaNet model architecture was used and modified in different ways to find the best performing model. The models contain parameters that are found during the training process but some parameters, called hyperparameters, need to be determined in advance. A way to automatically find good values of these hyperparameters was also tested. In hyperparameter optimization, the Bayesian optimization method proved to create a model with equally good performance as the best performance acieved by the author using manual hyperparameter tuning. The OD system was implemented using Keras with Tensorflow backend and received a high perfomance (mAP=0.9245) on the test data. The system manages to detect the wanted objects in the images but is expected to perform worse in a general situation since the training data and test data are very similar. In order to further develop this system and to improve performance under general conditions more data is needed from other airfields and under different weather conditions.

Infrared small target detection is widely used in fields such as military and security. UNet, which is a classical semantic segmentation method proposed in 2015, has shown excellent performance and robustness. However, U-Net suffers from the problem of losing small targets in deep layers after multiple down-sampling operations. Dilated convolution, as a special convolution that can increase the receptive field without increasing the number of parameters, is considered to be able to optimize the problems caused by down-sampling. Dense Nested Attention Network (DNANet), due to its superior performance, was chosen as the baseline, but it still has the issue of target loss. This study proposes three optimization directions: deep down-sampling replaced by cascaded dilated convolution, dilated spatial attention, and dilated residual block. In these three directions, this study proposes four methods, respectively DNANet-DS-1, DNANet-DS-2, DNANet-Att, and DNANet-RB. Two open-source infrared small target datasets, NUDT-SIRST and NUAA-SIRST, were used in this study. The four proposed methods were trained and tested on these two datasets. Among them, DNANetRB significantly outperforms other methods on the NUAA-SIRST dataset, so further experiments were conducted to observe the influence of different network depths on DNANet-RB. The experimental result indicates that when the network depth exceeds a certain threshold, the network can only achieve marginal improvements, but the number of parameters will increase significantly. / Infraröd detektering av små mål används ofta inom områden som militär och säkerhet. U-Net, som är en klassisk semantisk segmenteringsmetod som föreslogs 2015, har visat utmärkt prestanda och robusthet. U-Net lider dock av problemet med att förlora små mål i djupa lager efter flera nedprovningsoperationer. Dilaterad konvolution, som är en speciell konvolution som kan öka det receptiva fältet utan att öka antalet parametrar, anses kunna optimera de problem som orsakas av downsampling. DNANet (Dense Nested Attention Network) valdes som baslinje på grund av dess överlägsna prestanda, men det har fortfarande problemet med målförlust. Denna studie föreslår tre optimeringsriktningar: djup nedsampling ersatt av kaskad dilaterad konvolution, dilaterad rumslig uppmärksamhet och dilaterat restblock. I dessa tre riktningar föreslår denna studie fyra metoder, respektive DNANet-DS-1, DNANet-DS-2, DNANet-Att och DNANet-RB. Två dataset med små infraröda mål med öppen källkod, NUDT-SIRST och NUAA-SIRST, användes i denna studie. De fyra föreslagna metoderna tränades och testades på dessa två datamängder. Bland dem överträffar DNANet-RB betydligt andra metoder på NUAA-SIRST-datasetet, så ytterligare experiment genomfördes för att observera påverkan av olika nätverksdjup på DNANet-RB. Det experimentella resultatet visar att när nätverksdjupet överskrider ett visst tröskelvärde kan nätverket bara uppnå marginella förbättringar, men antalet parametrar kommer att öka avsevärt.

An image that has been subject to the out-of-focus phenomenon has reducedsharpness, contrast and level of detail depending on the amount of defocus. Torestore out-of-focused images is a complex task due to the information loss thatoccurs. However there exist many restoration algorithms that attempt to revertthis defocus by estimating a noise model and utilizing the point spread function.The purpose of this thesis, proposed by FLIR Systems, was to find a robustalgorithm that can restore focus and from the customer’s perspective be userfriendly. The thesis includes three implemented algorithms that have been com-pared to MATLABs built-in. Three image series were used to evaluate the limitsand performance of each algorithm, based on deblurring quality, implementationcomplexity, computation time and usability.Results show that the Alternating Direction Method for total variation de-convolution proposed by Tao et al. [29] together with its the modified discretecosines transform version restores the defocused images with the highest qual-ity. These two algorithms include features such as, fast computational time, fewparameters to tune and a powerful noise reduction.

This thesis presented an overview of the development of a geolocating algorithm as part of a geoprocessor for raw satellite imagery. This algorithm was devised for and limited by the specifications of a state-of-the-art multispectral telescope designed by Aistech Space, hosted onboard the Guardian spacecraft, which will observe Earth through the visible, near infrared, and thermal infrared bands of the electromagnetic spectrum. The geolocation algorithm presented here is composed of the combination of two models. The first is a physical model, which makes use of spacecraft telemetry and external satellite-tracking data to approximate the geographical center of a sensed scene. Secondly, an optical model obtains a reference Landsat image based on the timestamp and approximated location of the sensed scene and utilizes image processing techniques to pinpoint a more precise geographical location of the sensed scene within acceptable limits. This performance was achieved in 77% of the cases considered. To conclude, a roadmap of the subsequent development topics and their relevance was laid out. / Detta examensarbete presenterar en översikt för utvecklingen av en geolokaliseringsalgoritm som en del av en geoprocessor för obearbetade satellitbilder. Algoritmen anpassades för och begränsades av specifikationerna för ett toppmodernt multispektralt teleskop designat av Aistech Space. Teleskopet kommer att finnas ombord på rymdfarkosten Guardian, där den är avsedd att observera jorden i de synliga, nära infraröda och termiska infraröda delarna av det elektromagnetiska spektrumet. Geolokaliseringsalgoritmen som presenteras i detta arbete är sammansatt av en kombination av två modeller. Den första är en fysisk modell, vilken använder sig av rymdfarkostens telemetri och extern satellitspårningsdata för att approximera det geografiska centrumet av en plats. Den andra är en optisk modell, vilken använder sig av en Landsat-referensbild baserad på tidsstämpeln och den ungefärliga positionen av platsen och använder sedan bildbehandlingstekniker för att fastställa en mer exakt geografisk position av platsen inom acceptabla gränser. Denna prestation lyckades uppnås i 77% av de övervägda fallen. Avslutningsvis lades en plan ut för de efterföljande utvecklingsämnena och deras relevans.