De nombreux travaux de recherche sont actuellement menés afin de rendre les caméras infrarouges plus compactes et moins chères. En infrarouge refroidi, le défi est de proposer un système cryogénique compact pouvant être intégré sur un système à faible capacité d’emport tel qu’un drone. Dans ce cadre, l’utilisation d’une lentille mince en remplacement du filtre froid du cryostat permettrait de limiter la masse supplémentaire à refroidir et de maintenir constant le temps de descente en froid. En infrarouge non refroidi, l’objectif est de concevoir un petit capteur infrarouge bas coût « grand public » que l’on pourra inviter dans nos maisons, nos voitures, voire nos smartphones. L’utilisation d’une lentille mince ouvrirait la voie à des imageurs infrarouges peu onéreux.Dans ce contexte, j’ai choisi d’étudier le comportement d’une lentille de Fresnel dite d’ordre élevé intégrée dans une configuration optique de type landscape lens. J’ai montré que cette architecture optique mince peut fonctionner sur une large bande spectrale et sur un grand champ de vue. Cependant, les lentilles de Fresnel d’ordre élevé étant mal modélisées dans la littérature, j’ai développé mes propres algorithmes de modélisation afin de prévoir les performances d’un tel système. Grâce à cette étude, j’ai ensuite proposé deux systèmes d’imagerie, l’un refroidi et l’autre non refroidi. Chacun des deux systèmes a fait l’objet d’un prototype et a été entièrement caractérisé expérimentalement. Les résultats expérimentaux obtenus m’ont permis de valider les performances anticipées théoriquement et de mettre en évidence un phénomène de chromatisme diffractif latéral. Ces systèmes ouvrent la voie à deux nouvelles générations de caméras infrarouges. J’ai montré que l’imageur infrarouge refroidi possède une qualité image satisfaisante pour des applications d’aide au pilotage. Le prototype non refroidi est lui entièrement compatible avec des applications domotiques. Il a suscité l’intérêt de différents acteurs industriels. / Miniaturizing infrared optical systems is a research area of great interest nowadays in order to make them lighter and cheaper. In the cooled infrared domain, the objective is to design a compact cryogenic camera that could be integrated in a small-capacity carrier like a drone. To that purpose, replacing the cold filter of the dewar by a thin lens would limit the cooled down mass and would stabilize the cool down time. In the uncooled infrared domain, the objective is to design a small general use camera at a low cost. Its use could be generalized in houses, cars or even smartphones. The use of a thin lens would also pave the way for low-cost infrared imagers. In this context, I chose to study the imagery properties of a high order Fresnel lens integrated in a landscape lens architecture. I have demonstrated that this architecture can be used within a wide spectral range and over a wide field of view. However, current optical design software perform poorly on high order Fresnel lenses. Therefore, I have developed my own algorithms to model the performances of such a system. With that study, I have been able to design two prototypes with their own objectives: the first one is cooled and the second one is uncooled. Both systems have been demonstrated and entirely characterized. The experiment results have validated the theoretical performances of the systems and they highlighted an original kind of lateral chromatic aberration.These two systems pave the way to two new generations of infrared cameras. Indeed, on one hand I have proved that the cooled infrared quality may be good enough to qualify for an aircraft piloting aid. On the other hand, the uncooled prototype is fully compatible with low cost surveillance applications and the system raised the interest of various companies.

La possibilité d’obtenir une émission laser efficace sur une large bande spectrale autour de 1.05 µm a été démontrée récemment en utilisant des monocristaux CaF2 and SrF2 dopés Nd3+. Cette émission, éteinte du fait de la relaxation croisée entre ions dans les matériaux de type fluorite, peut être exaltée avec l’insertion d’ions “tampons” comme Y3+ ou Lu3+. Une émission laser accordable sur une large bande spectrale et la génération d’impulsions ultracourtes sont donc possibles. De plus, ce type de matériau est particulièrement intéressant pour des amplificateurs de forte puissance pompés par diodes. Une étude complète des propriétés spectrales et thermomécaniques de cristaux de CaF2:Nd3+ co-dopés avec des concentrations variables de Lu3+ a été menée en collaboration avec le CEA CESTA au sein du projet LASCAN pour de futurs développements au sein du Laser MégaJoule (LMJ) avec pour but l’amélioration des performances des amplificateurs pilotes. L’effet d’élimination des clusters de Nd3+ par le co-dopage avec le Lu3+ et l’apparition de deux sites actifs différents a été mise en évidence et caractérisée en détail. Ce travail présente ainsi une spectroscopie complète des centres actifs dans CaF2:Nd3+,Lu3+, une étude des propriétés thermomécaniques, ainsi que des propriétés d’amplification optique des matériaux étudiés. / It was recently demonstrated that efficient and broadband laser emission was possible with Nd-doped CaF2 and SrF2 single crystals around 1.05 µm. Such laser emission, known as completely quenched because of cross-relaxation in the singly doped fluorites, increases spectacularly by co-doping with non-optically active “buffer” ions like Y3+ or Lu3+. Broadband laser emission and ultra-short laser operation are therefore possible. The material is particularly appealing for large scale high peak power diode-pumped amplifiers. A deep investigation of the spectral properties of CaF2:Nd3+ crystals co-doped with various amounts of Lu3+ was performed. The cluster breaking effect of Lu3+ codoping has as a side-effect the appearance of two different contributions, identified as two different active sites, which is in agreement with recent observations. In this work, we investigate in depth the different centres’ spectroscopic properties and operational parameters for laser applications,their evolution as the lutetium concentration increases, and the thermomechanical properties of said samples. A preliminary investigation on the amplification properties and laser operation of said materials is performed as well.

Langmuir Circulation is a common phenomenon driven by wind in oceans and lakes and was first studied by Langmuir in 1927. According to various ocean observations, this kind of phenomenon plays an important role in many phenomena such as the aggregation of bubbles, the distribution of plankton as well as the mixing of spilled oil and sediment in the ocean. To study this, an experimental facility has been developed in the lab which creates a small scale version of Langmuir Circulation. This thesis is about the design and testing of this tank and surrounding aluminum frame, as well as the design and construction of the illumination equipment (the Green Lantern 2.0) needed for Particle Image Velocimetry measurements within the tank. ANSYS will be used to show whether the tank is structurally strong enough to support the fluid. An enhancement is found that prevents a frontward bend of tank wall, which is analyzed by ANSYS to find an optimized construction to minimize tank deformation. Then, the Light-Emitting Diode (LED) and collimating lens selection for the Green Lantern 2.0 will also be shown in this paper. Besides, this thesis also presents preliminary flow measurement data acquired using the illumination equipment (the Green Lantern).

This thesis aims to study how flat dielectric lenses can be designed. The usage of flat lenses is steadily increasing as they are smaller and less bulky than traditional convex lenses. Instead of a lens with a curved surface the permittivity in the lens is varied to achieve the same effect. Two different computational methods were investigated when approaching this problem: physical and geometrical optics. In physical optics the incoming radio waves are treated as waves in contrast to geometrical optics where it is considered as rays. Both methods are used as approximations of Maxwell's equations. The variation of permittivity in the lens was formulated as an optimisation problem where the lens' focusing abilities were maximised. The optimisation was implemented with an interior point method. Both arbitrary permittivity distributions as well as predetermined distributions were examined in this work. All optimised lens models were then simulated in a full wave commercial simulation software to verify and compare the two. The simulations showed that both approaches gave promising results as they focused the electromagnetic wave in a satisfying way. However the physical optics approach was more prominent as the focused radio waves had a much higher magnitude than the approach based on geometrical optics. The conclusion was therefore that physical optics is the preferred approach.

Femtosecond laser direct writing in transparent materials such as glass and optical fibers has been used as a versatile tool in order to fabricate various 3-D photonic structures such as active and passive waveguides, couplers, gratings and diffractive optical elements (DOEs). This capability of patterning and refractive index modification in the bulk of transparent materials depends on the nonlinear absorption phenomenon. This practical technique has the potential to be used for cost effective and simplified manufacturing in various applications. This thesis examines three advanced manufacturing techniques that use ultrashort pulse filamentary propagation induced by nonlinear absorption in the transparent materials. First, a new gradient index lens fabrication method using femtosecond laser direct writing is introduced. Light that passes through the lens with refractive index change resulting from localized energy deposition is focused using a beam profiler. Second, wide welding area of glass samples are used to fabricate microfluidic devices with long channels by adopting customized fixture. The fixture making artificial pressure helps the two glass samples have wide optical contact area and the highly intensive pulse filamentation strongly joins glass slides. As an example of a more specific application, microfluidic samples with long grooves sealed by femtosecond laser welding were successfully fabricated as part of this project. Finally, a screw-shaped, long period grating sensor was fabricated by rotating the optical fiber. This technique enables the fiber core to have asymmetric refractive index change, resulting in higher sensitivity compared to conventional long period grating sensors. Also, a new long-period grating sensor with reverse bending effect has been demonstrated by producing complex pitches of refractive index change. / Graduate

Indiana University-Purdue University Indianapolis (IUPUI) / Depth estimation is increasingly becoming more important in computer vision. The requirement for autonomous systems to gauge their surroundings is of the utmost importance in order to avoid obstacles, preventing damage to itself and/or other systems or people. Depth measuring/estimation systems that use multiple cameras from multiple views can be expensive and extremely complex. And as these autonomous systems decrease in size and available power, the supporting sensors required to estimate depth must also shrink in size and power consumption. This research will concentrate on a single passive method known as Depth from Defocus (DfD), which uses an in-focus and out-of-focus image to infer the depth of objects in a scene. The major contribution of this research is the introduction of a new Deep Learning (DL) architecture to process the the in-focus and out-of-focus images to produce a depth map for the scene improving both speed and performance over a range of lighting conditions. Compared to the previous state-of-the-art multi-label graph cuts algorithms applied to the synthetically blurred dataset the DfD-Net produced a 34.30% improvement in the average Normalized Root Mean Square Error (NRMSE). Similarly the DfD-Net architecture produced a 76.69% improvement in the average Normalized Mean Absolute Error (NMAE). Only the Structural Similarity Index (SSIM) had a small average decrease of 2.68% when compared to the graph cuts algorithm. This slight reduction in the SSIM value is a result of the SSIM metric penalizing images that appear to be noisy. In some instances the DfD-Net output is mottled, which is interpreted as noise by the SSIM metric. This research introduces two methods of deep learning architecture optimization. The first method employs the use of a variant of the Particle Swarm Optimization (PSO) algorithm to improve the performance of the DfD-Net architecture. The PSO algorithm was able to find a combination of the number of convolutional filters, the size of the filters, the activation layers used, the use of a batch normalization layer between filters and the size of the input image used during training to produce a network architecture that resulted in an average NRMSE that was approximately 6.25% better than the baseline DfD-Net average NRMSE. This optimized architecture also resulted in an average NMAE that was 5.25% better than the baseline DfD-Net average NMAE. Only the SSIM metric did not see a gain in performance, dropping by 0.26% when compared to the baseline DfD-Net average SSIM value. The second method illustrates the use of a Self Organizing Map clustering method to reduce the number convolutional filters in the DfD-Net to reduce the overall run time of the architecture while still retaining the network performance exhibited prior to the reduction. This method produces a reduced DfD-Net architecture that has a run time decrease of between 14.91% and 44.85% depending on the hardware architecture that is running the network. The final reduced DfD-Net resulted in a network architecture that had an overall decrease in the average NRMSE value of approximately 3.4% when compared to the baseline, unaltered DfD-Net, mean NRMSE value. The NMAE and the SSIM results for the reduced architecture were 0.65% and 0.13% below the baseline results respectively. This illustrates that reducing the network architecture complexity does not necessarily reduce the reduction in performance. Finally, this research introduced a new, real world dataset that was captured using a camera and a voltage controlled microfluidic lens to capture the visual data and a 2-D scanning LIDAR to capture the ground truth data. The visual data consists of images captured at seven different exposure times and 17 discrete voltage steps per exposure time. The objects in this dataset were divided into four repeating scene patterns in which the same surfaces were used. These scenes were located between 1.5 and 2.5 meters from the camera and LIDAR. This was done so any of the deep learning algorithms tested would see the same texture at multiple depths and multiple blurs. The DfD-Net architecture was employed in two separate tests using the real world dataset. The first test was the synthetic blurring of the real world dataset and assessing the performance of the DfD-Net trained on the Middlebury dataset. The results of the real world dataset for the scenes that were between 1.5 and 2.2 meters from the camera the DfD-Net trained on the Middlebury dataset produced an average NRMSE, NMAE and SSIM value that exceeded the test results of the DfD-Net tested on the Middlebury test set. The second test conducted was the training and testing solely on the real world dataset. Analysis of the camera and lens behavior led to an optimal lens voltage step configuration of 141 and 129. Using this configuration, training the DfD-Net resulted in an average NRMSE, NMAE and SSIM of 0.0660, 0.0517 and 0.8028 with a standard deviation of 0.0173, 0.0186 and 0.0641 respectively.

Imaging systems have long been designed in separated steps: the experience-driven optical design followed by sophisticated image processing. Such a general-propose approach achieves success in the past but left the question open for specific tasks and the best compromise between optics and post-processing, as well as minimizing costs. Driven from this, a series of works are proposed to bring the imaging system design into end-to-end fashion step by step, from joint optics design, point spread function (PSF) optimization, phase map optimization to a general end-to-end complex lens camera. To demonstrate the joint optics application with image recovery, we applied it to flat lens imaging with a large field of view (LFOV). In applying a super-resolution single-photon avalanche diode (SPAD) camera, the PSF encoded by diffractive op tical element (DOE) is optimized together with the post-processing, which brings the optics design into the end-to-end stage. Expanding to color imaging, optimizing PSF to achieve DOE fails to find the best compromise between different wavelengths. Snapshot HDR imaging is achieved by optimizing a phase map directly. All works are demonstrated with prototypes and experiments in the real world. To further compete for the blueprint of end-to-end camera design and break the limits of a simple wave optics model and a single lens surface. Finally, we propose a general end-to-end complex lens design framework enabled by a differentiable ray tracing image formation model. All works are demonstrated with prototypes and experiments in the real world. Our frameworks offer competitive alternatives for the design of modern imaging systems and several challenging imaging applications.

This thesis is focused on application development for simulation lenses’ optical distortions and also for creation own patterns. In the first part are discussed common problems of optical distortion and concept of software analysis. Further is described realization and implementation of particular modules in the application. In the end is show up graphical user interface and its functionality.

No description available.

No description available.