GPU-CPU implementation for super-resolution mosaicking of Unmanned Aircraft System (UAS) surveillance video

Camargo, Aldo; Schultz, Richard R.; Wang, Yi; Fevig, Ronald; He, Qiang

doi:10.1109/ssiai.2010.5483926

Cited by 13 publications

(11 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The performance experiment is conduct on task tree based mosaicking program TTLMosaic (TTM) and traditional MPI-implemented parallel mosaicking program (PM) [28] both with increasing computation resources and with increasing amount image data respectively. The performance experiment with increasing processors chose the case of mosaicking 200 15-meter panchromatic ETM+ [6] scenes.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Task-Tree Based Large-Scale Mosaicking for Massive Remote Sensed Imageries with Dynamic DAG Scheduling

Wang

Zomaya

et al. 2014

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

Remote sensed imagery mosaicking at large scale has been receiving increasing attentions in regional to global research. However, when scaling to large areas, image mosaicking becomes extremely challenging for the dependency relationships among a large collection of tasks which give rise to ordering constraint, the demand of significant processing capabilities and also the difficulties inherent in organizing these enormous tasks and RS image data. We propose a task-tree based mosaicking for remote sensed imageries at large scale with dynamic DAG scheduling. It expresses large scale mosaicking as a data-driven task tree with minimal height. And also a critical path based dynamical DAG scheduling solution with status queue named CPDS-SQ is provided to offer an optimized schedule on multi-core cluster with minimal completion time. All the individual dependent tasks are run by a core parallel mosaicking program implemented with MPI to perform mosaicking on different pairs of images. Eventually, an effective but easier approach is offered to improve the large-scale processing capability by decoupling the dependence relationships among tasks from the complex parallel processing procedure. Through experiments on large-scale mosaicking, we confirmed that our approach were efficient and scalable.

show abstract

Section: Methodsmentioning

confidence: 99%

“…In addition, GPUs with CUDA programming have become a competitive accelerator [5]. Samargo et al [28] presented CUDA accelerated mosaicking for unmanned aircraft system. The problem is that the complicated control logical of dependent tasks complicates the CUDA implementing.…”

Section: Related Workmentioning

confidence: 99%

Task-Tree Based Large-Scale Mosaicking for Massive Remote Sensed Imageries with Dynamic DAG Scheduling

Wang

Zomaya

et al. 2014

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

show abstract

“…In recent years, GPUs (Graphics Processing Units) with CUDA(Compute Unified Device Architecture) programming haveevolved into a highly parallel, multithreaded, many-core processors with tremendous computational speed and very high memory bandwidth [55]. Several relevant studies have illustrated GPU-based implementation of RS image processing [56], spatial data analysis [57]. In the field of parallel image mosaicking, Camargo et al [58] presented CUDA can significantly accelerated mosaicking for unmanned aircraft system.…”

Section: Image Mosaicking Implemented With Gpusmentioning

confidence: 99%

A review of parallel computing for large-scale remote sensing image mosaicking

Chen

Liu

et al. 2015

Cluster Comput

View full text Add to dashboard Cite

Interest in image mosaicking has been spurred by a wide variety of research and management needs. However, for large-scale applications, remote sensing image mosaicking usually requires significant computational capabilities. Several studies have attempted to apply parallel computing to improve image mosaicking algorithms and to speed up the calculation process. The state of art of this field has not yet been summarized, which is, however, essential for a better understanding and for further research of image mosaicking parallelism on a large scale. This paper provides a perspective on the current state of image mosaicking parallelization for large scale application. We firstly introduce the motivation of image mosaicking parallel for large scale application, and analysis the difficulty and problem of parallel image mosaicking at large scale such as scheduling with huge number of dependent tasks, programming with multiple-step procedure, dealing with frequent I/O operation.. Then the we summarize the current state of parallel computing in image mosaicking for large scale applications with respect to problem decomposition and parallel strategy, parallel architecture, task schedule strategy and implementation of image mosaicking parallelization. Finally, the key problems and future potential research directions for image mosaics are addressed.

show abstract

“…20,21 Capable of high performance computing, GPUs developed by NVIDIA are seeing increased use for signal and data processing in aerospace and defense applications. 22,23 Our short term solution for onboard video processing system includes a stripped down laptop with an Intel Corei7 CPU and a mobile kepler-class NVIDIA card.…”

Section: Onboard Video Processing Systemmentioning

confidence: 99%

Tracking of Mobile Targets using Unmanned Aerial Vehicles

Bhandari

Raheja

Rose

et al. 2012

AIAA Guidance, Navigation, and Control Conference

View full text Add to dashboard Cite

This paper presents an integrated effort by Cal Poly Pomona and UC Irvine for the tracking of mobile targets using unmanned aerial vehicles. A twin-engine airplane is used as the aerial platform. The airplane is being equipped with a commercial-off-the-shelf imaging and video processing systems, which are capable of tracking mobile targets. Besides using the off-the-shelf systems, we are also developing neuromorphic algorithms for autonomous target tracking and action selection. The algorithm is inspired by the mammalian visual system function and is implemented through GPU parallel computing. The algorithm uses a neuromorphic image preprocessing pipeline implemented on a NVIDIA GPU using CUDA and a visual tracking pipeline implemented on a CPU. For autonomous flight, the airplane is being equipped with a Piccolo II autopilot from Cloud Cap Technologies. Our UAV path planning strategy includes either circular or sinusoidal patterns, depending on the target speed. The path planning algorithm will be implemented on top of the autopilot system.

show abstract

GPU-CPU implementation for super-resolution mosaicking of Unmanned Aircraft System (UAS) surveillance video

Cited by 13 publications

References 8 publications

Task-Tree Based Large-Scale Mosaicking for Massive Remote Sensed Imageries with Dynamic DAG Scheduling

Task-Tree Based Large-Scale Mosaicking for Massive Remote Sensed Imageries with Dynamic DAG Scheduling

A review of parallel computing for large-scale remote sensing image mosaicking

Tracking of Mobile Targets using Unmanned Aerial Vehicles

Contact Info

Product

Resources

About