• Journal of the Korea Society of Computer and Information
• /
• 제27권10호
• /
• pp.11-17
• /
• 2022

In this paper, a model that can increase the efficiency of work in arranging interior furniture by applying augmented reality technology was studied. In the existing system to which augmented reality is currently applied, there is a problem in that information is limitedly provided depending on the size and nature of the company's product when outputting the image of furniture. To solve this problem, this paper presents an AR labeling algorithm. The AR labeling algorithm extracts feature points from the captured images and builds a database including indoor location information. A method of detecting and learning the location data of furniture in an indoor space was adopted using the CNN technique. Through the learned result, it is confirmed that the error between the indoor location and the location shown by learning can be significantly reduced. In addition, a study was conducted to allow users to easily place desired furniture through augmented reality by receiving detailed information about furniture along with accurate image extraction of furniture. As a result of the study, the accuracy and loss rate of the model were found to be 99% and 0.026, indicating the significance of this study by securing reliability. The results of this study are expected to satisfy consumers' satisfaction and purchase desires by accurately arranging desired furniture indoors through the design and implementation of AR labels.

• Journal of Korea Society of Industrial Information Systems
• /
• 제13권4호
• /
• pp.1-13
• /
• 2008

In this paper, we propose a method to solve a problem in binarizing the rear number plate image captured by a camera on a moving vehicle. An image may be shadowed by the cavernous structure of the rear side of a moving vehicle and it makes us hard to get a high quality of binary image. Therefore, we first detect a shadow edge and then divide an image into the shadow part and non-shadow part by the edge. Finally, the binary image is obtained by binarizing each part and merging them In this paper, we do comparative work on a group of binarization methods including our method, the method suggested by Zheng, the method using block binarization, and the method using labeling. The result shows that our method achieves better performance than others in most cases.

• Journal of the Korean Institute of Intelligent Systems
• /
• 제22권2호
• /
• pp.198-204
• /
• 2012

Tracking and recognition of robots are required for the cooperation task of robots in various environments. In the paper, a tracking control system of moving robot using stereo image processing, code-book model and fuzzy controller is proposed. First, foreground and background images are separated by using code-book model method. A candidate region is selected based on the color information in the separated foreground image and real distance of the robot is estimated from matching process of depth image that is acquired through stereo image processing. The open and close processing of image are applied and labeling according to the size of mobile robot is used to recognize the moving robot effectively. A fuzzy tracking controller using distance information and mobile information by stereo image processing is designed for effective tracking according to the movement velocity of the target robot. The proposed fuzzy tracking control method is verified through tracking experiments of mobile robots with stereo camera.

• The KIPS Transactions:PartB
• /
• 제16B권4호
• /
• pp.271-280
• /
• 2009

In this paper, we propose a smoke detection method using region growing method in outdoor video sequences. Our proposed method is composed of three steps; the initial change area detection step, the boundary finding and expanding step, and the smoke classification step. In the first step, we use a background subtraction to detect changed areas in the current input frame against the background image. In difference images of the background subtraction, we calculate a binary image using a threshold value and apply morphology operations to the binary image to remove noises. In the second step, we find boundaries of the changed areas using labeling algorithm and expand the boundaries to their neighbors using the region growing algorithm. In the final step, ellipses of the boundaries are estimated using moments. We classify whether the boundary is smoke by using the temporal information.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• 제5권6호
• /
• pp.1157-1162
• /
• 2001

This paper deals with inspect algorithm using visual system. One of the major problems that arise during polymer production is the estimation of the noise of the color product.(bad pallets) An erroneous output can cause a lot of losses (production and financial losses). Therefore new methods for real-time inspection of the noise are demanded. For this reason, we have presented a development of vision system algorithm for the defect inspection of PE color pallets. First of all, in order to detect the edge of object, the differential filter is used. And we apply to the labelling algorithm for feature extraction. This algorithm is designed for the defect inspection of pallets. The labelling algorithm permits to separate all of the connected components appearing on the pallets. Labelling the connected regions of a image is a fundamental computation in image analysis and machine vision, with a large number of application. Also, we suggested vision processing program in window environment. Simulations and experimental results demonstrate the performance of the proposal algorithm.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• 제65권11호
• /
• pp.1879-1886
• /
• 2016

In this paper, we propose a real-time recognition algorithm of lanes, stop lines and speed bumps on roads for autonomous vehicles. First, we generate a top-view using the image transmitted from a camera that is installed to see the front of a vehicle. To speed up the processing, we simplify the mapping algorithm in constructing a top-view wherein the region of interest (ROI) is concerned. The features of lanes, stop lines and speed bumps, which are composed of lines, are searched in the edge image of the top-view, then followed by labeling and clustering specialized to detect straight lines. The width of lines, distances from the center of a vehicle, and curvature of each cluster are considered to select final candidates. We verify the proposed algorithm on real roads using the commercial car (KIA K7) which is converted into an autonomous vehicle.

• Journal of the Korean Society of Industry Convergence
• /
• 제23권3호
• /
• pp.481-489
• /
• 2020

Safety and security during the handling of hazardous materials is a great concern for anyone in the field. One driving point in the security field is the ability to detect the source of the danger and take action against it as quickly as possible. Via the usage of a fully convolutional network, it is possible to create the label map of an input image, indicating what object is occupying the specific area of the image. This research employs the usage of U-net, which was constructed in biomedical field segmentation to segment cells, instead of the original FCN. One of the challenges that this research faces is the availability of ground truth with precise labeling for the dataset. Testing the network after training resulted in some images where the network pronounces even better detail than the expected label map. With better detailed label map, the network might be able to produce better segmentation is something to be studied in further research.

• IEIE Transactions on Smart Processing and Computing
• /
• 제3권4호
• /
• pp.195-204
• /
• 2014

This paper presents the depth evaluation for object detection by automated assembling robots. Pattern distortion analysis from a structured light system identifies an object with the greatest depth from its background. An automated assembling robot should prior select and pick an object with the greatest depth to reduce the physical harm during the picking action of the robot arm. Object detection is then combined with a depth evaluation to provide contour, showing the edges of an object with the greatest depth. The contour provides shape information to an automated assembling robot, which equips the laser based proxy sensor, for picking up and placing an object in the intended place. The depth evaluation process using structured light for an automated electronics assembling robot is accelerated for an image frame to be used for computation using the simplest experimental set, which consists of a single camera and projector. The experiments for the depth evaluation process required 31 ms to 32 ms, which were optimized for the robot vision system that equips a 30-frames-per-second camera.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권6호
• /
• pp.2302-2316
• /
• 2015

We propose an efficient method of extracting targets within a region of interest in non-homogeneous infrared images by using a principal component analysis (PCA) plane and adaptive Gaussian kernel. Existing approaches for extracting targets have been limited to using only the intensity values of the pixels in a target region. However, it is difficult to extract the target regions effectively because the intensity values of the target region are mixed with the background intensity values. To overcome this problem, we propose a novel PCA based approach consisting of three steps. In the first step, we apply a PCA technique minimizing the total least-square errors of an IR image. In the second step, we generate a binary image that consists of pixels with higher values than the plane, and then calculate the second derivative of the sum of the square errors (SDSSE). In the final step, an iteration is performed until the convergence criteria is met, including the SDSSE, angle and labeling value. Therefore, a Gaussian kernel is weighted in addition to the PCA plane with the non-removed data from the previous step. Experimental results show that the proposed method achieves better segmentation performance than the existing method.

• Journal of Korea Multimedia Society
• /
• 제24권8호
• /
• pp.1000-1011
• /
• 2021

In recent years, using Deep Learning methods to apply for medical and biomedical image analysis has seen many advancements. In clinical, using Deep Learning-based approaches for cancer image analysis is one of the key applications for cancer detection and treatment. However, the scarcity and shortage of labeling images make the task of cancer detection and analysis difficult to reach high accuracy. In 2015, the Unet model was introduced and gained much attention from researchers in the field. The success of Unet model is the ability to produce high accuracy with very few input images. Since the development of Unet, there are many variants and modifications of Unet related architecture. This paper proposes a new approach of using Unet++ with pretrained EfficientNet as backbone architecture for breast tumor cell nuclei segmentation and uses the multi-organ transfer learning approach to segment nuclei of breast tumor cells. We attempt to experiment and evaluate the performance of the network on the MonuSeg training dataset and Triple Negative Breast Cancer (TNBC) testing dataset, both are Hematoxylin and Eosin (H & E)-stained images. The results have shown that EfficientUnet++ architecture and the multi-organ transfer learning approach had outperformed other techniques and produced notable accuracy for breast tumor cell nuclei segmentation.