Guntae Bae

Abandoned luggage detection using a finite state automaton in surveillance video

We present an abandoned luggage detection architecture that consists of an intelligent surveillance system for public places. Detection of abandoned luggage is necessary because unattended or abandoned luggage can be used as a means of terrorist attack, especially for bombs. Our proposed system relies on three modules: moving object detection, object tracking and classification, and event recognition. We focus on abandoned luggage detection. To recognize an abandoned luggage event, we constructed the finite state automaton (FSA), in which each FSA state represents a certain luggage status. The proposed algorithm shows good performance in a real-world environment and also works at real-time speed.

Moving-object segmentation using a foreground history map

This paper describes a real-time foreground segmentation method in monocular video sequences for video teleconferencing. Background subtraction is widely used in foreground segmentation for static cameras. However, the results are usually not accurate enough for background substitution tasks. In this paper, we propose a novel strategy for fast and accurate foreground segmentation. The strategy consists of two steps: initial foreground segmentation and fine foreground segmentation. The key to our algorithm consists of two steps. In the first step, a moving object is roughly segmented using the background subtraction method. In order to update the initial foreground segmentation results in the second step, a region-based segmentation method and a foreground history map (FHM)-based segmentation representing the combination of temporal and spatial information were developed. The segmentation accuracy of the proposed algorithm was evaluated with respect to the ground truth, which was the manually cropped foreground. The experimental results showed that the proposed algorithm improved the accuracy of segmentation with respect to Horpraserts well-known algorithm.

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.

Unusual Event Recognition for Mobile Alarm System

This paper proposes an unusual event recognition algorithm, which is a part of a mobile alarm system. Our systems focus on unusual event. When the system detects the unusual event, the photos of emergency situation are passed to the users portable devices such as mobile phone or PDA along with event description to help the users final decision. The system combines the foreground segmentation, object tracking and unusual event recognition to detect the Drop off, Abandonand Steal bagevent. The event recognition module constructs the Bayesian network of each event and uses inference algorithm to detect the unusual event. The proposed system tested in PETS2006 and CAVIAR dataset. The proposed algorithm showed good results on the real world environment and also worked at real time speed.

An Illumination Invariant Traffic Sign Recognition in the Driving Environment for Intelligence Vehicles

This paper proposes a traffic sign recognition method in real road environments. The video stream in driving environments has two different characteristics compared to a general object video stream. First, the number of traffic sign types is limited and their shapes are mostly simple. Second, the camera cannot take clear pictures in the road scenes since there are many illumination changes and weather conditions are continuously changing. In this paper, we improve a modified census transform(MCT) to extract features effectively from the road scenes that have many illumination changes. The extracted features are collected by histograms and are transformed by the dense descriptors into very high dimensional vectors. Then, the high dimensional descriptors are encoded into a low dimensional feature vector by Fisher-vector coding and Gaussian Mixture Model. The proposed method shows illumination invariant detection and recognition, and the performance is sufficient to detect and recognize traffic signs in real-time with high accuracy.

Interactive optimization of photo composition with Gaussian mixture model on mobile platform

A good photo is determined using various visual elements of photography and these elements have been implemented in mobile devices with functionalities including zooming, auto-focusing and auto-white-balancing. Although composition is an important element of a good photo and an interesting research topic, most composition-related functionalities have not been added to mobile devices. We propose a guide system for capturing good photos in mobile devices that considers composition elements. A photo composition mixture model (PCMM) is derived based on composition elements such as a Gaussian Mixture Model (GMM), and the best composition of current input is gradually determined by iterating the PCMM optimization. Experimental evaluations are conducted to show the usefulness of the proposed PCMM and its optimization performance. To show the efficiency of recomposition performance and speed, we compare our method with retargeting-based methods. By implementing our method in mobile devices, we show that our system offers valid user guidance for capturing a photo with good composition in realtime.

Interactive Object Segmentation System from a Video Sequence

In this paper, we present an interactive object segmentation system form video, such as TV products and films, for converting 2D to 3D contents. It is focused on reducing the processing time for the object segmentation, increasing the usability. The proposed system is consist of three steps which are trimap generation based on polygon and object segmentation using Graph Cut algorithm and refinement by a user interfaces (UI) based on rectangle and local features. It makes it easy to get object segmentation rapidly. It is also helpful to create 3D contents.

Automatic background substitution using monocular camera and temporal foreground probability model

This paper describes an automatic background substitution for PC-softphone applications. Different from previous works on this topic which are mostly using stereo camera, we provide a solution which can segment the foreground region with monocular camera which is not moving. There are three main contributions from this paper. First, our proposed algorithm can segment the foreground region without any manual initialization. Second, we improve accuracy of the foreground segmentation using the Temporal Foreground Probability Model(TFPM). Finally, natural composite image was achieved by adaptive color and intensity correction method

Unusual behavior detection in the entry gate scenes of subway station using Bayesian networks and inference

In this paper, we propose a method for detecting unusual human behavior using monocular camera which is not moving. Our system composed of three modules which are moving object detection, tracking, and event recognition. The key part is event recognition module. We define unusual events which are composed of two simple events (drop off luggage, unattended luggage) and two complex events (abandoned luggage and steal luggage). In order to detect the simple event, we construct Bayesian network in each unusual event. We extract evidences using bounding box properties which are the location of moving objects, speed, distance between the person and the other moving object (such as bag), existing time. And then, we use finite state automaton which shows the temporal relation of two simple events to detect complex events. To evaluate the performance, we compare the frame number when an even is triggered with our results and the ground truth. The proposed algorithm showed good results on the real world environment and also worked at real time speed.