Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
2 results
Search Results
Item Restricted IN-LABORATORY SIMPLIFIED IMAGE-BASED EMPIRICAL POLARIMETRIC BIDIRECTIONAL REFLECTANCE DISTRIBUTION FUNCTION MEASUREMENT USING 3D GEOMETRIC TARGETS(University of Dayton, 2024-04-21) Aldkeelalah, Sultan; Ratliff, BradleyThe bidirectional reflectance distribution function (BRDF) is essential to remote sensing, computer graphics, and material science applications. It aids the development of photo-realistic rendering models, target detection and recognition, and atmospheric characterization tasks in remote sensing. BRDF measurements are cumbersome to make, requiring dense, full hemispherical sampling that often results in millions of individual measurements conducted at precise sensor and illumination source geometries. Methods have been proposed based upon theoretical, experimental, and empirical approaches that aim to simplify the data collection process. Empirically, researchers have proposed using targets of different geometric shapes in conjunction with image-based sensors to acquire measurements in parallel, thus reducing the acquisition time. This work surveys studies of such proposed methods to measure BRDF/polarimetric BRDF (pBRDF). Our group previously presented a fast and simple framework for empirically measuring the pBRDF using a linear imaging polarimeter from novel 3D-printed geodesic target spheres with well-characterized surface facets under outdoor environmental conditions. The models derived from this approach were validated against physics-based models and demonstrated good agreement. In this work, we present a modified approach to conduct similar measurements on the same faceted objects in a laboratory environment. The Applied Sensing Lab at the University of Dayton has constructed a solar simulation laboratory that allows for highly accurate and repeatable positioning of light sources, sensors, and objects. The laboratory contains both collimated (direct sun) and diffuse (downwelling) light sources that we have spectrally tuned in this work to match expected solar irradiance under a range of outdoor conditions. The laboratory-based pBRDF models obtained by our proposed framework validate strongly against their corresponding outdoor, spectroradiometric measurements (ground truth), and physics-based models.27 0Item Restricted Activation Functions In Deep Learning For Aerial Image Segmentation(Saudi Digital Library, 2023-11-01) Alamri, Raghad Jaza; Morley, TerenceIn remote sensing, deep learning models have been widely proposed and evaluated, especially for scene classification using Convolutional Neural Networks (CNNs) or semantic segmentation through Fully Convolutional Networks (FCN). There is still a research gap in studying the impact of activation functions on semantic segmentation performance in FCN, mainly when applied to aerial images. This dissertation attempts to bridge this gap by comprehensively examining the impact of nine activation func- tions on FCN models. This study presents intensive experiments on different FCN architectures, UNet and FPN. UNet is a simple and straightforward architecture, while FPN is very deep and complex. Also, two datasets were used: a small dataset with only five classes with images from the same country and a more diverse dataset with nine classes and images of various resolutions and complexity from all over the world. This experiment consists of two phases. The first phase involves establishing four baseline models for integrating diverse activation functions through a systematic method of hy- perparameter tuning. Afterwards, each baseline model was implemented across ten different activation function variations. In total, forty distinct models were trained and evaluated. Based on these experiments, it is evident that the choice of activation func- tions has a significant impact on the stability of the training and convergence speed. Additionally, the activation functions play a crucial role in the overall performance and within-class performance of the models. However, the behaviour of each activa- tion function is highly affected by the combination of architectures and datasets used.7 0