• IEMEK Journal of Embedded Systems and Applications
• /
• v.19 no.1
• /
• pp.9-17
• /
• 2024

It is widely recognized that for 3D human pose estimation (HPE), dataset acquisition is expensive and the effectiveness of augmentation techniques of conventional visual recognition tasks is limited. We address these difficulties by presenting a simple but effective method that augments input images in terms of viewpoints when training a 3D human pose estimation (HPE) model. Our intuition is that meaningful variants of the input images for HPE could be obtained by viewing a human instance in the images from an arbitrary viewpoint different from that in the original images. The core idea is to synthesize new images that have self-occlusion and thus are difficult to predict at different viewpoints even with the same pose of the original example. We incorporate this idea into the training procedure of the 3D HPE model as an augmentation stage of the input samples. We show that a strategy for augmenting the synthesized example should be carefully designed in terms of the frequency of performing the augmentation and the selection of viewpoints for synthesizing the samples. To this end, we propose a new metric to measure the prediction difficulty of input images for 3D HPE in terms of the distance between corresponding keypoints on both sides of a human body. Extensive exploration of the space of augmentation probability choices and example selection according to the proposed distance metric leads to a performance gain of up to 6.2% on Human3.6M, the well-known pose estimation dataset.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.5
• /
• pp.971-981
• /
• 2012

Google map has become one of the most recognized and comfortable means for providing statistical information of geographically referenced data. In this article, we introduce R functions to embed google map images on R interface and develop a function to represent statistical graphs such as bar graph, pie chart, and rectangle graph on a google map images.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.7
• /
• pp.76-90
• /
• 1994

In this paper, an efficient algorithm is proposed which recognizes the mixed document consisting of the printed Korean/alphanumeric texts and graphic images. In the preprocessing step an input document is aligned if necessary by rotating it. We obtain the rotation angle using the Hough transform and align the input document horizontally. Then we separate graphic image parts from text parts by considering chain codes of connected components. We further separate each character using vertical and horizontal projections. In the recognition step Korean and alphanumeric characters are classified and each of them is recognized hierarchically using several features. In summary an efficient recognition algorithm for mixed documents is proposed and its performance is demonstrated via computer simulations.

• Journal of Multimedia Information System
• /
• v.2 no.1
• /
• pp.171-178
• /
• 2015

This article describes 3 dimensional (3D) face recognition system using histogram of oriented gradients (HOG) based on face curvature. The surface curvatures in the face contain the most important personal feature information. In this paper, 3D face images are recognized by the face components: cheek, eyes, mouth, and nose. For the proposed approach, the first step uses the face curvatures which present the facial features for 3D face images, after normalization using the singular value decomposition (SVD). Fisherface method is then applied to each component curvature face. The reason for adapting the Fisherface method maintains the surface attribute for the face curvature, even though it can generate reduced image dimension. And histogram of oriented gradients (HOG) descriptor is one of the state-of-art methods which have been shown to significantly outperform the existing feature set for several objects detection and recognition. In the last step, the linear discriminant analysis is explained for each component. The experimental results showed that the proposed approach leads to higher detection accuracy rate than other methods.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.49 no.6
• /
• pp.307-315
• /
• 2000

For the autonomy of a mobile robot it is first needed to know its position and orientation. Various methods of estimating the position of a robot have been developed. However, it is still difficult to localize the robot without any initial position or orientation. In this paper we present the method how to make the colored map and how to calculate the position and direction of a robot using the angle data of an omnidirectional image. The wall of the map is rendered with the corresponding color images and the color histograms of images and the coordinates of feature points are stored in the map. Then a mobile robot gets the color omnidirectional image at arbitrary position and orientation, segments it and recognizes objects by multiple color indexing. Using the information of recognized objects robot can have enough feature points and localize itself.

• Korean Journal of Computational Design and Engineering
• /
• v.16 no.5
• /
• pp.331-343
• /
• 2011

High tibial osteotomy (HTO) is a widely accepted treatment for unicompartmental osteoarthritis of the knee and other lower extremity deformities, particularly for young and active patients. However, it is generally recognized as a technically demanding procedure. Thus the limitations of conventional surgical methods have been reported. Currently, the use of computer-aided surgery (CAS) system for successful surgery was increased. However it has been reported many problems such as expensive equipment, infection or fracture caused by inserting half-pin and delaying surgery. Therefore we propose a novel method which can be tracked by using tracking-pin inserted in arbitrary locations of the femur and tibia, and fluoroscopic images obtained from G-arm of antero-posterior and lateral-medial.

• The Bulletin of The Korean Astronomical Society
• /
• v.41 no.2
• /
• pp.67.2-67.2
• /
• 2016

The Sun has diverse variations in solar atmosphere's layers due to solar activity. This solar variations can be recognized easily by sunspots which appear on the solar photosphere. Thus the sunspot on the photosphere is utilized by direct index of the solar activity. The other variation of the photosphere is center-to-limb variation (CLV). In this study, we analyze the relative intensity observed by SOHO, SDO. The data of photospheric intensity are from full disk images of SOHO/MDI intensity ($6768{\AA}$, from May 1994 to March 2011) and of SDO/HMI intensity ($6173-6174{\AA}$, from May 2010 to June 2016). As the result, we found the latitudinal variation of the intensity. The daily photospheric intensity showed the solar cyclic variation with sunspot number. It has a little difference of phase with sunspot number.

• Journal of the Korean Data and Information Science Society
• /
• v.18 no.4
• /
• pp.1033-1043
• /
• 2007

The multi layer perceptron for segmenting objects in images only uses the input windows that are made from a image in a fixed size. These windows are recognized so each independent learning data that they make the performance of the multi layer perceptron poor. The poor performance is caused by not considering the position information and effect of input windows in input images. So we propose a new approach to add the position information and effect of input windows to the multi layer perceptron#s input layer. Our new approach improves the performance as well as the learning time in the multi layer perceptron. In our experiment, we can find our new algorithm good.

• Journal of Korea Society of Industrial Information Systems
• /
• v.5 no.1
• /
• pp.53-60
• /
• 2000

In this paper, we developed a character recognition system which could be applied to the automation of construction assembling, and testing process of components by the recognition of characters on the components used in industry department. In this system the image information of each component was caught by the CCD camera and then characters on the images were recognized automatically. If we apply the system to the industrial field it will meet the various requirements such as high productivity, low cost and high qualify products and factory automation.

• Journal of the Semiconductor & Display Technology
• /
• v.15 no.2
• /
• pp.43-48
• /
• 2016

In this paper, we propose a traffic light detection and recognition (TLDR) algorithm in the daytime. The proposed algorithm utilizes the color and shape information for the TLDR. At first, a traffic light is detected and recognized based on its shape information. Then, the color range of the detected traffic light is investigated in HSV color space. The input data of the proposed TLDR algorithm is the color image captured using the black box camera during driving. Our simulations demonstrate that the proposed algorithm can achieve a high detection and recognition performance for the images including traffic lights.