• Journal of Broadcast Engineering
• /
• v.22 no.4
• /
• pp.509-518
• /
• 2017

High-resolution digital holograms can be quickly generated by using a PC cluster that is based on server-client architecture and is composed of several GPU-equipped PCs. However, the data transmission time between PCs becomes a large obstacle for fast generation of video holograms because it linearly increases in proportion to the number of frames. To resolve the problem with the increase of data transmission time, this paper proposes a multi-threading-based method. Hologram generation in each client PC basically consists of three processes: acquisition of light sources, CGH operation using GPUs, and transmission of the result to the server PC. Unlike the previous method that sequentially executes the processes, the proposed method executes in parallel them by multi-threading and thus can significantly reduce the proportion of the data transmission time to the total hologram generation time. Through experiments, it was confirmed that the total generation time of a high-resolution video hologram with 150 frames can be reduced by about 30%.

• Journal of KIISE:Computer Systems and Theory
• /
• v.34 no.11
• /
• pp.572-580
• /
• 2007

Wireless sensor networks are sensing, computing and communication infrastructures that allow us to monitor, instrument, observe, and respond to phenomena in the harsh environment. Generally, the wireless sensor networks are composed of many deployed sensor nodes that were designed to be very cost-efficient in terms of production cost. For example, UC Berkeley's MICA motes have only 8-bit CPU, 4KB RAM, and 128KB FLASH memory space. Therefore, sensor operating systems that run on the sensor nodes should be able to operate efficiently in terms of the resource management. In this paper, we present a dynamic threads stack management scheme for space-constrained and multi-threaded sensor operating systems. In this scheme, the necessary stack space of each function is measured on compile-time. Then, the information is used to dynamically allocate and release each function's stack space on run-time. It was implemented in Nano-Qplus sensor operating system. Our experimental results show that the proposed scheme outperforms the existing fixed-size stack allocation mechanism.

• The KIPS Transactions:PartA
• /
• v.18A no.1
• /
• pp.19-24
• /
• 2011

As era of multi-core processors has arrived, transactional memory has been considered as an effective method to achieve easy and fast multi-threaded programming. Various hardware transactional memory systems such as UTM, VTM, FastTM, LogTM, and LogTM-SE, have been introduced in order to implement high-performance multi-core processors. Especially, LogTM-SE has provided study performance with an efficient memory management policy and a practical thread scheduling method through conflict detection based on signatures. However, increasing number of cores on a processor imposes the hardware complexity for signature processing. This causes overall performance degradation due to the heavy workload on signature comparison. In this paper, we propose a new architecture of multiple signature comparison to improve conflict detection of signature based transactional memory systems.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.35 no.3
• /
• pp.187-194
• /
• 2017

The present study introduces a method to parallelize k-means clustering algorithm for fast unsupervised classification of large satellite imagery. Known as a representative algorithm for unsupervised classification, k-means clustering is usually applied to a preprocessing step before supervised classification, but can show the evident advantages of parallel processing due to its high computational intensity and less human intervention. Parallel processing codes are developed by using multi-threading based on OpenMP. In experiments, a PC of 8 multi-core integrated CPU is involved. A 7 band and 30m resolution image from LANDSAT 8 OLI and a 8 band and 10m resolution image from Sentinel-2A are tested. Parallel processing has shown 6 time faster speed than sequential processing when using 10 classes. To check the consistency of parallel and sequential processing, centers, numbers of classified pixels of classes, classified images are mutually compared, resulting in the same results. The present study is meaningful because it has proved that performance of large satellite processing can be significantly improved by using parallel processing. And it is also revealed that it easy to implement parallel processing by using multi-threading based on OpenMP but it should be carefully designed to control the occurrence of false sharing.

• Journal of Internet Computing and Services
• /
• v.19 no.1
• /
• pp.97-103
• /
• 2018

In oriental paintings, there is Luo-kuan that expressed in a single picture by compressing the artist's information. Such Luo-kuan includes various information such as the title of the work or the name of the artist. Therefore, information about Luo-kuan is considered important to those who collect or enjoy oriental paintings. However, most of the letters in the Luo-kuan are difficult kanji, kanzai, or various shapes, so it is difficult for the ordinary people to interpret. In this paper, we developed an Luo-kuan search application to easily check the information of the Luo-kuan. The application uses a search algorithm that analyzes the captured Luo-kuan image and sends it to the server to output information about the Luo-kuan candidates that are most similar to the Luo-kuan images taken from the database in the server. We also compared and analyzed the accuracy of the algorithm based on 170 Luo-kuan data in order to find out the ranking of the Luo-kuan that matched the Luo-kuan among the candidates. Accuracy Analysis Experimental Results The accuracy of the search algorithm of this application is confirmed to be about 90%, and it is anticipated that it will be possible to develop a platform to automatically analyze and search images in a big data environment by supplementing the optimizing algorithm and multi-threading algorithm.