• The Journal of the Korea institute of electronic communication sciences
• /
• v.15 no.3
• /
• pp.563-572
• /
• 2020

Tidal flats provide important indicators that inform the condition of the environment, so we need to monitor them systematically. Currently, the projects to monitor tidal flats by periodically observing the creatures in tidal flats are underway. Still, it is done in a way that people observe directly, so it is not systematic and efficient. In this paper, we propose a background modeling method for tidal flat images that can be applied to a system that automatically monitors creatures living in tidal flats using sensor network technology. The application of sensor network technology makes it difficult to collect enough images due to the limitation of transmission capacity. Therefore, in this paper, we propose a method to effectively model the background and generate foreground maps by reflecting the characteristics of tidal flat images in the situation where the number of images to be used for analysis is small. Experimental results show that the proposed method models the background of a tidal flat image easily and accurately.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.6
• /
• pp.227-233
• /
• 2019

Recently, along with the recent development of deep learning technique, neural networks are achieving success in computer vision filed. Convolutional neural network have shown outstanding performance in not only for a simple image classification task, but also for tasks with high difficulty such as object segmentation and detection. However many such deep learning models are based on supervised-learning, which requires more annotation labels than image-level label. Especially image semantic segmentation model requires pixel-level annotations for training, which is very. To solve these problems, this paper proposes a weakly-supervised semantic segmentation method which requires only image level label to train network. Existing weakly-supervised learning methods have limitations in detecting only specific area of object. In this paper, on the other hand, we use multi-classifier deep learning architecture so that our model recognizes more different parts of objects. The proposed method is evaluated using VOC 2012 validation dataset.

• Journal of Broadcast Engineering
• /
• v.16 no.1
• /
• pp.44-53
• /
• 2011

In this paper, we propose a 3D conversion algorithm of 2D video using depth layer partition method. In the proposed algorithm, we first set frame groups using cut detection algorithm. Each divided frame groups will reduce the possibility of error propagation in the process of motion estimation. Depth image generation is the core technique in 2D/3D conversion algorithm. Therefore, we use two depth map generation algorithms. In the first, segmentation and motion information are used, and in the other, edge directional histogram is used. After applying depth layer partition algorithm which separates objects(foreground) and the background from the original image, the extracted two depth maps are properly merged. Through experiments, we verify that the proposed algorithm generates reliable depth map and good conversion results.

• Journal of Broadcast Engineering
• /
• v.18 no.4
• /
• pp.643-646
• /
• 2013

We present an improved AdaBoost algorithm to avoid overfitting phenomenon. AdaBoost is widely known as one of the best solutions for object detection. However, AdaBoost tends to be overfitting when a training dataset has noisy samples. To avoid the overfitting phenomenon of AdaBoost, the proposed method divides positive samples into K clusters using k-means algorithm, and then uses only one cluster to minimize the training error at each iteration of weak learning. Through this, excessive partitions of samples are prevented. Also, noisy samples are excluded for the training of weak learners so that the overfitting phenomenon is effectively reduced. In our experiment, the proposed method shows better classification and generalization ability than conventional boosting algorithms with various real world datasets.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.39 no.2
• /
• pp.235-240
• /
• 2015

Active Safety system is requested on the market of the medium and heavy duty commercial vehicle over 4.5ton beside the market of passenger car with advancement of the digital equipment proportionally. Unlike the passenger car, the mounting position of camera in case of the medium and heavy duty commercial vehicle is relatively high, it is disadvantaged conditions for lane recognition in contradiction to passenger car. In this work, we show the method of lane recognition through the Sobel edge, based on the spatial domain processing, Hough transform and color conversion correction. Also we suggest the low error method of front vehicles recognition in order to reduce the detection error through Haar-like, Adaboost, SVM and Template matching, etc., which are the object recognition methods by frontal camera vision. It is verified that the reliability over 98% on lane recognition is obtained through the vehicle test.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.4
• /
• pp.95-104
• /
• 2012

In the image captured by the camera in a moving vehicle, there exist various motions due to the change of background, motion of objects in the image that make difficult to extract a pure vibrational motion by the camera. In this paper, we suggest an image stabilization with the elimination of various motion components based on the classification of motions in the image by their characteristics. After the elimination of various local motions, images are compensated and stabilized with the global motion caused by the camera. Also, we suggest an accurate and fast image stabilization by excluding regions of little information based on block differences and edge densities.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.4
• /
• pp.682-691
• /
• 2017

In this paper, we propose the fusion design methodology of both pedestrian detection and object tracking system realized with the aid of HOG-PCA based RBFNN pattern classifier. The proposed system includes detection and tracking parts. In the detection part, HOG features are extracted from input images for pedestrian detection. Dimension reduction is also dealt with in order to improve detection performance as well as processing speed by using PCA which is known as a typical dimension reduction method. The reduced features can be used as the input of the FCM-based RBFNNs pattern classifier to carry out the pedestrian detection. FCM-based RBFNNs pattern classifier consists of condition, conclusion, and inference parts. FCM clustering algorithm is used as the activation function of hidden layer. In the conclusion part of network, polynomial functions such as constant, linear, quadratic and modified quadratic are regarded as connection weights and their coefficients of polynomial function are estimated by LSE-based learning. In the tracking part, object tracking algorithms such as mean shift(MS) and cam shift(CS) leads to trace one of the pedestrian candidates nominated in the detection part. Finally, INRIA person database is used in order to evaluate the performance of the pedestrian detection of the proposed system while MIT pedestrian video as well as indoor and outdoor videos obtained from IC&CI laboratory in Suwon University are exploited to evaluate the performance of tracking.

• Journal of KIISE
• /
• v.45 no.1
• /
• pp.15-21
• /
• 2018

Low-Rank Representation (LRR) based methods are widely used in many practical applications, such as face clustering and object detection, because they can guarantee high prediction accuracy when used to constructing graphs in graph - based semi-supervised learning. However, in order to solve the LRR problem, it is necessary to perform singular value decomposition on the square matrix of the number of data points for each iteration of the algorithm; hence the calculation is inefficient. To solve this problem, we propose an improved and faster LRR method based on the recently published Fast LRR (FaLRR) and suggests ways to introduce and optimize additional constraints on the underlying optimization goals in order to address the fact that the FaLRR is fast but actually poor in classification problems. Our experiments confirm that the proposed method finds a better solution than LRR does. We also propose Fast MLRR (FaMLRR), which shows better results when the goal of minimizing is added.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.12
• /
• pp.2665-2674
• /
• 2012

This paper propose the method to recognize the lanes and distance between cars in real-time which detects dangerous situations and helps safe driving in the actual road environment. First of all, it extracts the area of interest corresponding to roads and cars from the road image photographed by using the forward-looking camera. Through the hough transform for the area of interest, this study detects linear components and also selects the lane and conducts filtering by calculating probability. And through the shadow threshold analysis of the cars in front within the area of interest, it extracts the objects of cars in front and calculates the distance from cars in front. According to the result of applying the suggested technology to recognize the lane and distance between cars to the road situation for testing, it showed over 95% recognition rate; thus, it has been proved that it can respond to safe driving.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.6
• /
• pp.87-97
• /
• 2015

An Aerial video summarization is not only the key to effective browsing video within a limited time, but also an embedded cue to efficiently congregative situation awareness acquired by unmanned aerial vehicle. Different with previous works, we utilize sensor operation mode of unmanned aerial vehicle, which is global, local, and focused surveillance mode in order for accurately summarizing the aerial video considering flight and surveillance/reconnaissance environments. In focused mode, we propose the moving-react tracking method which utilizes the partitioning motion vector and spatiotemporal saliency map to detect and track the interest moving object continuously. In our simulation result, the key frames are correctly detected for aerial video summarization according to the sensor operation mode of aerial vehicle and finally, we verify the efficiency of video summarization using the proposed mothed.