Advancements in digital camera technologies have led to significantly smaller camera sensors without compromising imaging quality. This trend opens avenues for replacing single-camera systems with multi-camera arrays, offering benefits such as an expanded field-of-view and 3D information. However, camera arrays require additional processing to interpret the collected image data. This report addresses two significant algorithmic challenges in developing a camera array. The first challenge involves virtual view generation, which is the task of estimating the image for a virtual camera using image data from multiview cameras. Ideally, if the geometry of the underlying scene is known, the virtual view can be generated by projecting the 3D geometry to the virtual camera. However, finding the 3D geometry of a scene from multiview cameras is an active research topic in Computer Vision. Considering the real-time processing requirement (<33 ms), we propose a real-time 3D visualization (RT3DV) system using a multiview RGB camera array that can process multiple synchronized video streams to produce a stereo video of a dynamic scene from a chosen view angle. We implemented a proof of concept RT3DV system tasked to process five synchronous video streams acquired by an RGB camera array. It achieves a processing speed of 44 milliseconds per frame and a peak signal-to-noise ratio (PSNR) of 15.9 dB from a viewpoint coinciding with a reference view. As a comparison, an image-based MVS algorithm will require 7 seconds to render a frame and yield a reference view PSNR of 16.3 dB. The second challenge is to stabilize a video at a real-time rate. We propose LSstab, a novel algorithm that efficiently suppresses unwanted motion jitters in real-time. LSstab features a parallel realization of the \textit{a-contrario} RANSAC (AC-RANSAC) algorithm to estimate the inter-frame camera motion parameters. A novel least-squares smoothing cost function is proposed to mitigate undesirable camera jitters. A recursive least square solver is then derived to minimize the smoothing cost function with a linear computation complexity. Evaluation against state-of-the-art video stabilization methods using publicly available videos demonstrates that LSstab achieves comparable or superior performance, particularly attaining real-time processing speed with GPU utilization.