• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.7C
• /
• pp.554-564
• /
• 2012

The vehicle recognition consists of two steps; the vehicle region detection step and the vehicle identification step based on the feature extracted from the detected region. Features using linear transformations have the effect of dimension reduction as well as represent statistical characteristics, and show the robustness in translation and rotation of objects. Among the linear transformations, the NMF(Non-negative Matrix Factorization) is one of part-based representation. Therefore, we can extract NMF features with sparsity and improve the vehicle recognition rate by the representation of local features of a car as a basis vector. In this paper, we propose a feature extraction using NMF suitable for the vehicle recognition, and verify the recognition rate with it. Also, we compared the vehicle recognition rate for the occluded area using the SNMF(sparse NMF) which has basis vectors with constraint and LVQ2 neural network. We showed that the feature through the proposed NMF is robust in the urban scene where occlusions are frequently occur.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.3
• /
• pp.31-39
• /
• 2006

This paper presents a novel approach to real-time recognition of 3D environment and objects for various applications such as intelligent robots, intelligent vehicles, intelligent buildings,..etc. First, we establish the three fundamental principles that humans use for recognizing and interacting with the environment. These principles have led to the development of an integrated approach to real-time 3D recognition and modeling, as follows: 1) It starts with a rapid but approximate characterization of the geometric configuration of workspace by identifying global plane features. 2) It quickly recognizes known objects in environment and replaces them by their models in database based on 3D registration. 3) It models the geometric details the geometric details on the fly adaptively to the need of the given task based on a multi-resolution octree representation. SIFT features with their 3D position data, referred to here as stereo-sis SIFT, are used extensively, together with point clouds, for fast extraction of global plane features, for fast recognition of objects, for fast registration of scenes, as well as for overcoming incomplete and noisy nature of point clouds.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.2
• /
• pp.751-770
• /
• 2019

Action recognition is an essential task in computer vision due to the variety of prospective applications, such as security surveillance, machine learning, and human-computer interaction. The availability of more video data than ever before and the lofty performance of deep convolutional neural networks also make it essential for action recognition in video. Unfortunately, limited crafted video features and the scarcity of benchmark datasets make it challenging to address the multi-person action recognition task in video data. In this work, we propose a deep convolutional neural network-based Effective Hybrid Learning (EHL) framework for two-person interaction classification in video data. Our approach exploits a pre-trained network model (the VGG16 from the University of Oxford Visual Geometry Group) and extends the Faster R-CNN (region-based convolutional neural network a state-of-the-art detector for image classification). We broaden a semi-supervised learning method combined with an active learning method to improve overall performance. Numerous types of two-person interactions exist in the real world, which makes this a challenging task. In our experiment, we consider a limited number of actions, such as hugging, fighting, linking arms, talking, and kidnapping in two environment such simple and complex. We show that our trained model with an active semi-supervised learning architecture gradually improves the performance. In a simple environment using an Intelligent Technology Laboratory (ITLab) dataset from Inha University, performance increased to 95.6% accuracy, and in a complex environment, performance reached 81% accuracy. Our method reduces data-labeling time, compared to supervised learning methods, for the ITLab dataset. We also conduct extensive experiment on Human Action Recognition benchmarks such as UT-Interaction dataset, HMDB51 dataset and obtain better performance than state-of-the-art approaches.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.9
• /
• pp.3762-3781
• /
• 2020

Infrared pedestrian target recognition is a vital problem of significant interest in computer vision. In this work, a novel infrared pedestrian target recognition method that uses heterogeneous features with multi-kernel transfer learning is proposed. Firstly, to exploit the characteristics of infrared pedestrian targets fully, a novel multi-scale monogenic filtering-based completed local binary pattern descriptor, referred to as MSMF-CLBP, is designed to extract the texture information, and then an improved histogram of oriented gradient-fisher vector descriptor, referred to as HOG-FV, is proposed to extract the shape information. Second, to enrich the semantic content of feature expression, these two heterogeneous features are integrated to get more complete representation for infrared pedestrian targets. Third, to overcome the defects, such as poor generalization, scarcity of tagged infrared samples, distributional and semantic deviations between the training and testing samples, of the state-of-the-art classifiers, an effective multi-kernel transfer learning classifier called MK-TrAdaBoost is designed. Experimental results show that the proposed method outperforms many state-of-the-art recognition approaches for infrared pedestrian targets.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1857-1862
• /
• 2016

Human activity recognition using depth information is an emerging and challenging technology in computer vision due to its considerable attention by many practical applications such as smart home/office system, personal health care and 3D video games. This paper presents a novel framework of 3D human body detection, tracking and recognition from depth video sequences using spatiotemporal features and modified HMM. To detect human silhouette, raw depth data is examined to extract human silhouette by considering spatial continuity and constraints of human motion information. While, frame differentiation is used to track human movements. Features extraction mechanism consists of spatial depth shape features and temporal joints features are used to improve classification performance. Both of these features are fused together to recognize different activities using the modified hidden Markov model (M-HMM). The proposed approach is evaluated on two challenging depth video datasets. Moreover, our system has significant abilities to handle subject's body parts rotation and body parts missing which provide major contributions in human activity recognition.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.493-496
• /
• 2009

Many AR (Augmented Reality) applications have been interested in a marker-less tracking since the tracking methods give camera poses without attaching explicit markers. In this paper, we propose a new marker-less page recognition and tracking algorithm for an AR book application such as DigilogBook. The proposed method only requires orthogonal images of pages, which need not to be trained for a long time, and the algorithm works in real-time. The page recognition is done in two steps by using SIFT (Scale Invariant Feature Transform) descriptors and the comparison evaluation function. And also, the method provides real-time tracking with 25fps ~ 30fps by separating the page recognition and the frame-to-frame matching into two multi-cores. The proposed algorithm will be extended to various AR applications that require multiple objects tracking.

• Journal of Institute of Control, Robotics and Systems
• /
• v.8 no.7
• /
• pp.599-606
• /
• 2002

A new method to govern the navigation of a mobile robot using hand gesture recognition is proposed based on the following two procedures. One is to achieve vision information by using a 2-DOF camera as a communicating medium between a man and a mobile robot and the other is to analyze and to control the mobile robot according to the recognized hand gesture commands. In the previous researches, mobile robots are passively to move through landmarks, beacons, etc. In this paper, to incorporate various changes of situation, a new control system that manages the dynamical navigation of mobile robot is proposed. Moreover, without any generally used expensive equipments or complex algorithms for hand gesture recognition, a reliable hand gesture recognition system is efficiently implemented to convey the human commands to the mobile robot with a few constraints.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.4 no.3
• /
• pp.197-201
• /
• 2011

The use of gestures provides an attractive alternate to cumbersome interfaces for human-computer devices interaction. This has motivated a very active research area concerned with computer vision-based recognition of hand gestures. The most important issues in hand gesture recognition is to recognize the directivity of finger. The primitive elements extracted to a hand gesture include in very important information on the directivity of finger. In this paper, we propose the recognition algorithm of finger directivity by using the cross points of circle and sub-primitive element. The radius of circle is increased from minimum radius including main-primitive element to it including sub-primitive elements. Through the experiment, we demonstrated the efficiency of proposed algorithm.

• Journal of Korea Multimedia Society
• /
• v.13 no.3
• /
• pp.400-409
• /
• 2010

This paper presents a position error recognition system when the wafer is mounted in cleaning equipment among the wafer manufacturing processes. The proposed system is to enhance the performance in cost and reliability by preventing the wafer cleaning system from damaging by alerting it when it is put in correct position. The key algorithms are the calibration method between image acquired from camera and physical wafer, a infrared lighting and the design of the filter, and the extraction of wafer boundary and the position error recognition resulting from generation of circle based on least square method. The system is to install in-line process using high reliable and high accurate position recognition. The experimental results show that the performance is good in detecting errors within tolerance.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.4
• /
• pp.699-708
• /
• 2021

In this paper, a steering control and speed control algorithm was presented for autonomous driving based on two vision sensors and road speed sign board. A car speed control algorithm was developed to recognize the speed sign by using TensorFlow, a deep learning program provided by Google to the road speed sign image provided from vision sensor B, and then let the car follows the recognized speed. At the same time, a steering angle control algorithm that detects lanes by analyzing road images transmitted from vision sensor A in real time, calculates steering angles, controls the front axle through PWM control, and allows the vehicle to track the lane. To verify the effectiveness of the proposed algorithm's steering and speed control algorithms, a car's prototype based on the Python language, Raspberry Pi and OpenCV was made. In addition, accuracy could be confirmed by verifying various scenarios related to steering and speed control on the test produced track.