• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.11
• /
• pp.4426-4442
• /
• 2020

Since scene segmentation is becoming a hot topic in the field of autonomous driving and medical image analysis, researchers are actively trying new methods to improve segmentation accuracy. At present, the main issues in image semantic segmentation are intra-class inconsistency and inter-class indistinction. From our analysis, the lack of global information as well as macroscopic discrimination on the object are the two main reasons. In this paper, we propose a Densely connected residual Attention network (DA-Res2Net) which consists of a dense residual network and channel attention guidance module to deal with these problems and improve the accuracy of image segmentation. Specifically, in order to make the extracted features equipped with stronger multi-scale characteristics, a densely connected residual network is proposed as a feature extractor. Furthermore, to improve the representativeness of each channel feature, we design a Channel-Attention-Guide module to make the model focusing on the high-level semantic features and low-level location features simultaneously. Experimental results show that the method achieves significant performance on various datasets. Compared to other state-of-the-art methods, the proposed method reaches the mean IOU accuracy of 83.2% on PASCAL VOC 2012 and 79.7% on Cityscapes dataset, respectively.

• International Journal of Internet, Broadcasting and Communication
• /
• v.13 no.1
• /
• pp.54-60
• /
• 2021

AR (Augmented Reality) is a technology that provides virtual content to the real world and provides additional information to objects in real-time through 3D content. In the past, a high-performance device was required to experience AR, but it was possible to implement AR more easily by improving mobile performance and mounting various sensors such as ToF (Time-of-Flight). Also, the importance of mobile augmented reality is growing with the commercialization of high-speed wireless Internet such as 5G. Thus, this paper proposes a system that can provide AR services via GNN (Graph Neural Network) using cameras and sensors on mobile devices. ToF of mobile devices is used to capture depth maps. A 3D point cloud was created using RGB images to distinguish specific colors of objects. Point clouds created with RGB images and Depth Map perform downsampling for smooth communication between mobile and server. Point clouds sent to the server are used for 3D object detection. The detection process determines the class of objects and uses one point in the 3D bounding box as an anchor point. AR contents are provided through app and web through class and anchor of the detected object.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.15 no.2
• /
• pp.46-56
• /
• 2012

Recently, when software architecture is designed, the efficiency of reusability is emphasized. The reusability of the design can raise the quality of GIS software, and reduce the cost of maintenance. Because the object oriented GoF design pattern provides the class hierarchy that can represent repetitively, the importance is emphasized more. This method that designs the GIS software can be applied from various application systems. A multiple distributed sensor network system is composed of the complex structure that each node of the sensor network nodes has different functions and sensor nodes and server are designed by the combination of many classes. Furthermore, this sensor network system may be changed into more complex systems according to a particular purpose of software designer. This paper will design the CATL model by applying Facade pattern that can enhance the efficiency of reuse according to attributes and behaviors in classes in order to implement the complicated structure of the multiple distributed sensor network system based on TinyOS. Therefore, our object oriented GIS design pattern model will be expected to utilize efficiently for design, update, or maintenance, etc. of new systems by packing up attributes and behaviors of classes at complex sensor network systems.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.12B
• /
• pp.2233-2248
• /
• 1999

The TMN that appears to operate the various communication networks generally and efficiently is developed under the different platform environment such as the different hardware and the different operating system. One of the main problems is that all the agents of the TMN system must be duplicated and maintain the software and the data blocks that perform the identical function. Therefore, the standard of the Q3 interface development cannot be defined and the multi-platform cannot be supported in the development of the TMN agent. In order to overcome these problems, the Farming methodology that is based on the Farmer model has been suggested. With the Farming methodology, the software and the data components which are duplicated and stored in each distributed object are saved in the platform independent class repository(PICR) by converting into the format of the independent componentware in the platform, so that the componentwares that are essential for the execution can be loaded and used statically or dynamically from PICR as described in the framework of each distributed object. The distributed TMN agent of the personal communication network is designed and developed by using the Farmer model.

• Journal of Information Processing Systems
• /
• v.17 no.4
• /
• pp.787-800
• /
• 2021

Three-dimensional (3D) object classification tasks using point clouds are widely used in 3D modeling, face recognition, and robotic missions. However, processing raw point clouds directly is problematic for a traditional convolutional network due to the irregular data format of point clouds. This paper proposes a pointwise convolution neural network (CNN) structure that can process point cloud data directly without preprocessing. First, a 2D convolutional layer is introduced to percept coordinate information of each point. Then, multiple 2D convolutional layers and a global max pooling layer are applied to extract global features. Finally, based on the extracted features, fully connected layers predict the class labels of objects. We evaluated the proposed pointwise CNN structure on the ModelNet10 dataset. The proposed structure obtained higher accuracy compared to the existing methods. Experiments using the ModelNet10 dataset also prove that the difference in the point number of point clouds does not significantly influence on the proposed pointwise CNN structure.

• International Journal of Computer Science & Network Security
• /
• v.21 no.7
• /
• pp.284-296
• /
• 2021

Quality Models are important element of the software industry to develop and implement the best quality product in the market. This type of model provides aid in describing quality measures, which directly enhance the user satisfaction and software quality. In software development, the inheritance technique is an important mechanism used in object-oriented programming that allows the developers to define new classes having all the properties of super class. This technique supports the hierarchy design for classes and makes an "is-a" association among the super and subclasses. This paper describes a standard procedure for validating the inheritance metric in Quality Model for Object-Oriented Design (QMOOD) by using a set of nine properties established by Weyuker. These properties commonly using for investigating the effectiveness of the metric. The integration of two measuring methods (i.e. QMOOD and Weyuker) will provide new way for evaluating the software quality based on the inheritance context. The output of this research shows the extent of satisfaction of the inheritance metric in QMOOD against Weyuker nine properties. Further results proved that Weyker's property number nine could not fulfilled by any inheritance metrics. This research introduces a way for measuring software that developed using object-oriented approach. The theoretical validation of the inheritance metric presented in this paper is a small step taken towards producing quality software and in providing assistance to the software industry.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10c
• /
• pp.310-312
• /
• 2003

망 운용에 있어서 망 관리 시스템은 다양한 네트워크 장비의 증가로 인하여 점점 더 중요한 위치를 차지하고 있다. 과거 망을 구성하는 요소는 네트워크 장비로 국한되었지만 현재는 서비스를 지원하는 데몬(Daemon) 등 장비 외적인 요소까지 속하게 되어 더욱 복잡해 지고 있다. 기존의 망 관리 시스템은 SNMP(Simple Network Management Protocol)나ICMP(Internet Control Message Protocol) 프로토콜을 사용하였으나 이 기술만으로는 데몬과 같은 장비 외적인 요소들까지 제어, 관리하지 못했다. 반면. TELNET등과 같은 기능을 추가적으로 이용할 수 있다면. 보다 쉽게 망 정보를 얻을 수 있는 장점이 있어 기존의 망 관리 시스템에서 지원하지 못했던 기능의 한계점을 극복할 수 있다. 이에 본 논문은 객체지향기술(OOT : Object­Oriented Technology)을 도입하였다. 클래스(Class)는 고유한 특성을 가지는 관리 대상 종류를 선언한 것이고. 관리 대상은 네트워크 장비 뿐만 아니라 데몬이나 지역, 부서등과 같은 추상적인 대상이 포함되며 이는 객체(Object)를 의미한다. 또한. 상속(Inheritance) 개념을 도입하여 중복요소를 없앴으며, 클래스의 메소드(Method)에 관리 대상을 처리할 수 있는 기능을 정의하였다. 기능 정의에는 SNMP와 TELNET. ICMP등을 내부적으로 이용하여 구현된 스크립트(Script)를 이용하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.12
• /
• pp.6009-6027
• /
• 2019

Support vector machine (SVM) classifiers have been widely used for object detection. These methods usually locate the object by finding the region with maximal score in an image. With bag-of-features representation, the SVM score of an image region can be written as the sum of its inside feature-weights. As a result, the searching process can be executed efficiently by using strategies such as branch-and-bound. However, the feature-weight derived by optimizing region classification cannot really reveal the category knowledge of a feature-point, which could cause bad localization. In this paper, we represent a region in an image by a collection of local feature-points and determine the object by the region with the maximum posterior probability of belonging to the object class. Based on the Bayes' theorem and Naive-Bayes assumptions, the posterior probability is reformulated as the sum of feature-scores. The feature-score is manifested in the form of the logarithm of a probability ratio. Instead of estimating the numerator and denominator probabilities separately, we readily employ the density ratio estimation techniques directly, and overcome the above limitation. Experiments on a car dataset and PASCAL VOC 2007 dataset validated the effectiveness of our method compared to the baselines. In addition, the performance can be further improved by taking advantage of the recently developed deep convolutional neural network features.

• Journal of the Korean Society for Precision Engineering
• /
• v.30 no.9
• /
• pp.901-908
• /
• 2013

The introduction of industrial intelligent robot using vision sensor has been interested in automated factory. 2D and 3D vision sensors have used to recognize object and to estimate object pose, which is for packaging parts onto a complete whole. But it is not trivial task due to illumination and various types of objects. Object image has distorted due to illumination that has caused low reliability in recognition. In this paper, recognition method of complex shape object has been proposed. An accurate object region has detected from combined binary image, which has achieved using DoG filter and local adaptive binarization. The object has recognized using neural network, which is trained with sub-divided object class according to object type and rotation angle. Predefined shape model of object and maximal slope have used to estimate the pose of object. The performance has evaluated on ETRI database and recognition rate of 96% has obtained.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.5
• /
• pp.1246-1256
• /
• 1998

MIB(Management Information Base), one of the key components of network management system, is a conceptual repository for the information of the various managed objects. MIB stores and manages all the structural and operational data of each managed resources. Therefore, MIB should be protected properly from inadvertant user access or malicious attacks. International standard ISO/IEC 10164-9 describes several managed object classes for the enforcement of MIB security. Those managed object classes described access control rules for security policy. But the exact authorization procedures using those newly added managed object classes are not presented. In this paper, we divide managed object classes into two groups, explicit and implicit ones, and describe the access authorization procedure in Z specification language. Using Z as a description method for both authorization procedure and GDMO's action part, the behaviour of each managed object class and access authorization procedure is more precisely and formally defined than those of natural language form.