• KIPS Transactions on Computer and Communication Systems
• /
• v.5 no.9
• /
• pp.237-246
• /
• 2016

DIBR (Depth Image Based Rendering) is a multimedia technology that generates the virtual multi-view images using a color image and a depth image, and it is used for creating glasses-less 3-dimensional display contents. This research describes the effect of depth accuracy about the objective quality of DIBR-based multi-view images. It first evaluated the minimum depth quantization bit that enables the minimum distortion so that people cannot recognize the quality degradation. It then presented the comparative analysis of non-uniform domain-division quantization versus regular linear quantization to find out how effectively express the accuracy of the depth information in same quantization levels according to scene properties.

• Journal of Korean Society for Geospatial Information Science
• /
• v.15 no.4
• /
• pp.111-116
• /
• 2007

This study is aimed to analyze the accuracy of depth information of reservoir using the robot-ship equipped with GPS and echosounder. The accuracy of depth measurements by sounding data was analyzed according to change of robot-ship's speed in the water. The field experiment results showed that as robot-ship's speeds were slow, accuracy of sounding data were increased. Until Robot-ship's speed was up to 5 km/hr, the accuracy of sounding data were included reliable section of normal distribution.

• Journal of Multimedia Information System
• /
• v.4 no.1
• /
• pp.43-50
• /
• 2017

In this paper, we propose a method which detects the nose and face of certain human by using the depth image. The proposed method has advantages of the low computational complexity and the high accuracy even in dark environment. Also, the detection accuracy of nose and face does not change in various postures. The proposed method first locates the locally protruding part from the depth image of the human body captured through the depth camera, and then confirms the nose through the depth characteristic of the nose and surrounding pixels. After finding the correct pixel of the nose, we determine the region of interest centered on the nose. In this case, the size of the region of interest is variable depending on the depth value of the nose. Then, face region can be found by performing binarization using the depth histogram in the region of interest. The proposed method can detect the nose and the face accurately regardless of the pose or the illumination of the captured area.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.22 no.3
• /
• pp.408-413
• /
• 2013

Micro-features such as grooves and lenses, which perform optical functions in flat displays, should be manufactured with a good form accuracy because this is directly related to their optical performance. As the size of the display increases, it is very difficult to maintain a high relative accuracy because of the inherent geometric errors such as the waviness of a large-area plate. In this paper, the optical effect of these geometric errors is investigated, and surface-referenced micro-grooving to measure and compensate for such geometric errors on line is proposed to improve the form accuracy of the micro-grooves. A PZT-based fast depth adjustment servo system is implemented in the tool holder to maintain a uniform groove depth in reference to the wavy surface. Through experiments, the proposed method is shown to be an efficient way to produce high-quality micro- grooves on a wavy die surface.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.252-255
• /
• 2020

Unsupervised deep learning methods have shown impressive results for the challenging monocular depth estimation task, a field of study that has gained attention in recent years. A common approach for this task is to train a deep convolutional neural network (DCNN) via an image synthesis sub-task, where additional views are utilized during training to minimize a photometric reconstruction error. Previous unsupervised depth estimation networks are trained within a fixed depth estimation range, irrespective of its possible range for a given image, leading to suboptimal estimates. To overcome this suboptimal limitation, we first propose an unsupervised adaptive depth estimation method guided by minimum and maximum (min-max) depth priors for a given input image. The incorporation of min-max depth priors can drastically reduce the depth estimation complexity and produce depth estimates with higher accuracy. Moreover, we propose a novel network architecture for adaptive depth estimation, called the AdaMM-DepthNet, which adopts the min-max depth estimation in its front side. Intensive experimental results demonstrate that the adaptive depth estimation can significantly boost up the accuracy with a fewer number of parameters over the conventional approaches with a fixed minimum and maximum depth range.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.17 no.6
• /
• pp.61-67
• /
• 2018

The face milling cutter, which is mainly used for the face milling, is used to cut the Carbon steel(SM20C) in the machining center for 5 times and 10 times respectively. This study clarify the dimensional accuracy characteristics according to the number of fine machining varied the condition of cutting depth, table feed speed and spindle speed. Cutting depth is varied 0.05~0.2mm, table feed speed is varied 0.05~0.2mm/min and spindle speed is varied 1500~2500rpm. As a result, the dimensional accuracy was stable 6 times machining with table feed speed 150mm/min and 10 times machining with table speed 100mm/min and cutting depth 0.05mm regardless times of machining.

• Journal of Korean Port Research
• /
• v.15 no.1
• /
• pp.75-85
• /
• 2001

Accuracy of dredging processes depends on the types of equipment used, the sediments encountered, whether the work to be performed is new or maintenance dredging, pre- and post-hydrographic surveying and so forth. Among those, position surveying accuracy which is directly determined by the control of the dredge's position and depth surveying accuracy being surveyed at the dredging point during dredging work are important factors. The purpose of this study is to develop 'Dredge Management System'for Grab dredge which is composed of 4 sub-system using LADGPS for dredge position determining system and dredging point determining system, tide gauge system and optical sensor for depth determining system and GIS and ENC for total management system. This system is installed on the grab dredge 'EUNJIN G-18'and applied to anchorage dredging work. at Pohang Harbor. The results revealed that this system is easy to operate, achieves good accuracy with only 45cm unevenness, reduces working period by 22 percent and saves cost 16.6 percent.

• Journal of the Korean Society for Precision Engineering
• /
• v.26 no.9
• /
• pp.53-57
• /
• 2009

There are technical requirements to manufacture large size functional parts with not only simple geometries like a flat or spherical surface but also sculptured geometries. In addition, the required machining accuracy for these parts is becoming more severe. In general, the form accuracy of machined parts is determined by the relative position between workpiece and tool during machining process. To improve machining accuracy the relative position errors should be maintained within the required accuracy. This study deals with the estimation and calibration of depth of cut using the AE signal in micro-machining. Also, this sensing technique can be applied to detect the initial contact between workpiece and tool.

• ETRI Journal
• /
• v.45 no.4
• /
• pp.666-677
• /
• 2023

In the last few years, convolutional neural networks (CNNs) have demonstrated good performance while solving various computer vision problems. However, since CNNs exhibit high computational complexity, signal processing is performed on the server side. To reduce the computational complexity of CNNs for edge computing, a lightweight algorithm, such as a MobileNet, is proposed. Although MobileNet is lighter than other CNN models, it commonly achieves lower classification accuracy. Hence, to find a balance between complexity and accuracy, additional hyperparameters for adjusting the size of the model have recently been proposed. However, significantly increasing the number of parameters makes models dense and unsuitable for devices with limited computational resources. In this study, we propose a novel MobileNet architecture, in which the number of parameters is adaptively increased according to the importance of feature maps. We show that our proposed network achieves better classification accuracy with fewer parameters than the conventional MobileNet.

• Journal of the Korean Society for Precision Engineering
• /
• v.16 no.4 s.97
• /
• pp.220-228
• /
• 1999

The machining accuracy has been improved with the development of NC machine tools and cutting tools. However, it is difficult to obtain a high degree of accuracy when machining deep pocket with long end mill, since machining accuracy is mainly dependant on the stiffness of the cutting tool. To improve surface accuracy in machining deep pocket using end mill, the performance by down cut and up cut is compared theoretically and experimentally. To verify usefulness of up milling, various experiments were carried out. As a result, it is found that up milling produce more accurate surface than down milling in machining deep pocket. For effective application of up milling, various values in helix angle, number of teeth, radial depth of cut and axial depth of cut are applied in experiment.