Sooyeong Kwak

Survey of computer vision–based natural disaster warning systems

With the rapid development of information technology, natural disaster prevention is growing as a new research field dealing with surveil- lance systems. To forecast and prevent the damage caused by natural disasters, the development of systems to analyze natural disasters using remote sensing geographic information systems (GIS), and vision sensors has been receiving widespread interest over the last decade. This paper provides an up-to-date review of five different types of natural disasters and their corresponding warning systems using computer vision and pattern recognition techniques such as wildfire smoke and flame detection, water level detection for flood prevention, coastal zone monitor- ing, and landslide detection. Finally, we conclude with some thoughts about future research directions.

Human interaction recognition in YouTube videos

This paper introduces the use of annotation tags for human activity recognition in video. Recent methods in human activity recognition use more complex and realistic datasets obtained from TV shows or movies, which makes it difficult to obtain the high recognition accuracies. We improve the recognition accuracies using annotation tags of the video. Tags tend to be related to video contents, and human activity videos frequently contain tags relevant to their activities. We first collect a human activity dataset containing tags from YouTube. Under this dataset, we automatically discover relevant tags and their correlation with human activities. We finally develop a framework using visual content and tags for activity recognition. We show that our approach can improve recognition accuracies compared with other approaches that only use visual content.

Detection of Abnormal Behavior by Scene Analysis in Surveillance Video

In intelligent surveillance system, various methods for detecting abnormal behavior were proposed recently. However, most researches are not robust enough to be utilized for actual reality which often has occlusions because of assumption the researches have that individual objects can be tracked. This paper presents a novel method to detect abnormal behavior by analysing major motion of the scene for complex environment in which object tracking cannot work. First, we generate Visual Word and Visual Document from motion information extracted from input video and process them through LDA(Latent Dirichlet Allocation) algorithm which is one of document analysis technique to obtain major motion information(location, magnitude, direction, distribution) of the scene. Using acquired information, we compare similarity between motion appeared in input video and analysed major motion in order to detect motions which does not match to major motions as abnormal behavior.

A Pedestrian Collision Warning System using a Fuzzy Logic

A pedestrian collision warning system which makes a judgement of pedestrians intention to help avoiding hitting accidents is proposed. This system uses the image sequences obtained from a car black box as well as vehicles speed obtained from a GPS. It detects pedestrians, if any, based on the Histogram of Gradient method and extracts several information such as the pedestrians relative positions, the direction of motion vectors, and distance between vehicle and pedestrian . A fuzzy logic based on these extracted information is applied to analyze the pedestrians safety levels. When the safety level is determined to be danger, an alarm is triggered to the driver. The performance of the proposed algorithm is tested under various driving scenarios, which shows it works successfully in real-time.

Golf Swing Classification Using Fuzzy System

A method to classify a golf swing motion into 7 sections using a Kinect sensor and a fuzzy system is proposed. The inputs to the fuzzy logic are the positions of golf club and its head, which are extracted from the information of golfer`s joint position and color information obtained by a Kinect sensor. The proposed method consists of three modules: one for extracting the joint`s information, another for detecting and tracking of a golf club, and the other for classifying golf swing motions. The first module extracts the hand`s position among the joint information provided by a Kinect sensor. The second module detects the golf club as well as its head with the Hough line transform based on the hand`s coordinate. Using a fuzzy logic as a classification engine reduces recognition errors and, consequently, improves the performance of robust classification. From the experiments of real-time video clips, the proposed method shows the reliability of classification by 85.2%.

Violent Behavior Detection using Motion Analysis in Surveillance Video

The demand of violence detection techniques using a video analysis to help prevent crimes is increasing recently. Many researchers have studied vision based behavior recognition but, violent behavior analysis techniques usually focus on violent scenes in television and movie content. Many methods previously published usually used both a color(e.g., skin and blood) and motion information for detecting violent scenes because violences usually involve blood scenes in movies. However, color information (e.g., blood scenes) may not be useful cues for violence detection in surveillance videos, because they are rarely taken in real world situations. In this paper, we propose a method of violent behavior detection in surveillance videos using motion vectors such as flow vector magnitudes and changes in direction except the color information. In order to evaluate the proposed algorithm, we test both USI dataset and various real world surveillance videos from YouTube.

3D mesh and multi-view synthesis implementation using stereo cameras and a depth camera

In this paper, we propose a new method of 3D mesh generation and view synthesis for multi-view video generation. The multi-view video can be produced by a view synthesis method and 3D mesh using RGB+Z images which are composed of one or two RGB images and one depth image. We create 36-view images using the view synthesis method and generate 3D mesh with vertexes and faces for free view rendering. Using the image synthesis method, warping, edge compensation, upsampling and 36-view generation processes are used. For 3D mesh, depth and RGB images are converted to the component of 3D mesh data. The quality of synthesized images is measured by MOS(Mean opinion score).