• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.2A
• /
• pp.157-164
• /
• 2002

The Mellin Transform is a well-known method to find out transformations between two overlapping images. This paper models a scene with an object moving with different velocities. Also, it analyzes the relationship of the detected correlation peak to the spectrum of the background and an object. Lastly, we investigate the optimal solution of the image registration parameters minimizing the effect of the noise from the Mellin Transform by applying the Wiener filter concept.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2021.06a
• /
• pp.36-38
• /
• 2021

본 논문은 물체 검출(Object Detection)과 물체영역분할(Object Segmentation)의 CNN 추론 결과를 MPEG-7 서술자 이진화를 통해 표현함으로써 원본과의 용량을 비교한다. 영상의 사용 목적에 따라 CNN 추론 결과를 압축하여 활용할 시 원본 영상 대비 용량을 측정하여 그 효율성을 판단하는 것이 목표이다. 물체 검출과 물체영역분할에 대한 추론 결과를 MPEG-7 서술자를 이용해 압축하였으며, 비교를 위해 원본 영상, CNN 추론 결과 파일, MPEG-7 서술자, MPEG-7 서술자 이진화 파일의 크기를 측정하였다. 실험 결과, MPEG-7 서술자를 이진화를 통한 표현 방식이 원본 영상 및 추론 결과 파일에 비해 효율적임을 알 수 있었다.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.3
• /
• pp.18-29
• /
• 2012

Viola-Jones algorithm, a very effective real-time object detection method, uses Haar-like features to constitute weak classifiers. A Haar-like feature is made up of at least two rectangles each of which corresponds to either positive or negative areas and the feature value is computed by subtracting the sum of pixel values in the negative area from that of pixel values in the positive area. Compared to the conventional Haar-like feature which is made up of more than one rectangle, in this paper, we present a couple of new rectangular features whose feature values are computed either by the sum or by the variance of pixel values in a rectangle. By the use of these rectangular features in combination with the conventional Haar-like features, we can select additional features which have been excluded in the conventional Viola-Jones algorithm where every features are the combination of contiguous bright and dark areas of an object. In doing so, we can enhance the performance of object detection without any computational overhead.

• Journal of KIISE
• /
• v.44 no.7
• /
• pp.668-673
• /
• 2017

Deep neural networks have shown remarkable performance in various fields of pattern recognition such as voice recognition, image recognition and object detection. However, underlying mechanisms of the network have not been fully revealed. In this paper, we focused on empirical analysis of the network parameters. The Faster R-CNN(region-based convolutional neural network) was used as a baseline network of our work and three important parameters were analyzed: the dropout ratio which prevents the overfitting of the neural network, the size of the anchor boxes and the activation function. We also compared the performance of dropout and batch normalization. The network performed favorably when the dropout ratio was 0.3 and the size of the anchor box had not shown notable relation to the performance of the network. The result showed that batch normalization can't entirely substitute the dropout method. The used leaky ReLU(rectified linear unit) with a negative domain slope of 0.02 showed comparably good performance.

• Journal of Broadcast Engineering
• /
• v.4 no.1
• /
• pp.22-31
• /
• 1999

Moving object tracking improves the efficiency and qualification for observation system, broadcasting system, video conference, etc. This paper propcses an improved Bayes decision method for detecting and tracking moving objects in active camera environment. The Bayes decision based tracking approach finds the region of moving objects by analyzing the image sequences statistically. The propcsed algorithm regenerates the probability density function to accord with moving objects and background for active camera. Experimental results show that the algorithm is accurate. reliable and noise resistant. The result is compared with those of the conventional methods.

• The KIPS Transactions:PartB
• /
• v.12B no.4 s.100
• /
• pp.403-412
• /
• 2005

In this paper, we propose an effective method for detecting and classifying shot transitions in video sequences. The proposed method detects and classifies shot transitions including cuts, fades and dissolves by compensating camera operations in video sequences, so that our method prevents false positives resulting from camera operations. Also, our method eliminates local moving objects in the process of compensating camera operations, so that our method prevents errors resulting from moving objects. In the experiments, we show that our shot transition approach can work as a promising solution by comparing the proposed method with previously known methods in terms of performance.

• The KIPS Transactions:PartB
• /
• v.11B no.3
• /
• pp.267-274
• /
• 2004

In this paper, the interesting moving objects tracking algorithm using color information on Multi-Video camera is proposed Moving objects are detected by using difference image method and integral projection method to background image and objects image only with hue area, after converting RGB color coordination of image which is input from multi-video camera into HSI color coordination. Hue information of the detected moving area are normalized by 24 steps from 0$^{\circ}$ to 360$^{\circ}$ It is used for the feature parameters of the moving objects that three normalization levels with the highest distribution and distance among three normalization levels after obtaining a hue distribution chart of the normalized moving objects. Moving objects identity among four cameras is distinguished with distribution of three normalization levels and distance among three normalization levels, and then the moving objects are tracked and surveilled. To examine propriety of the proposed method, four cameras are set up indoor difference places, humans are targeted for moving objects. As surveillance results of the interesting human, hue distribution chart variation of the detected Interesting human at each camera in under 10%, and it is confirmed that the interesting human is tracked and surveilled by using feature parameters at four cameras, automatically.

• The KIPS Transactions:PartB
• /
• v.9B no.1
• /
• pp.99-104
• /
• 2002

Existing methods of object tracking use template matching, re-detection of object boundaries or motion information. The template matching method requires very long computation time. The re-detection of object boundaries may produce false edges. The method using motion information shows poor tracking performance in moving camera. In this paper, a robust object tracking algorithm is proposed, using projected motion and histogram intersection. The initial object image is constructed by selecting the regions of interest after image segmentation. From the selected object, the approximate displacement of the object is computed by using 1-dimensional intensity projection in horizontal and vortical direction. Based on the estimated displacement, various template masks are constructed for possible orientations and scales of the object. The best template is selected by using the modified histogram intersection method. The robustness of the proposed tracking algorithm has been verified by experimental results.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.2C
• /
• pp.181-187
• /
• 2012

Millimeter wave (MMW) readily penetrates fabrics, thus it can be used to detect objects concealed under clothing. A passive MMW imaging system can operate as a stand-off type sensor that scans people in both indoors and outdoors. However, because of the diffraction limit and low signal level, the imaging system often suffers from low image quality. Therefore, suitable statistical analysis and computational processing would be required for automatic analysis of the images. In this paper, a real-time concealed object detection is addressed by means of the multi-level segmentation. The histogram of the image is modeled with a Gaussian mixture distribution, and hidden object areas are segmented by a multi-level scheme involving $k$-means, the expectation-maximization algorithm, and a decision rule. The complete algorithm has been implemented in C++ environments on a standard computer for a real-time process. Experimental and simulation results confirm that the implemented system can achieve the real-time detection of concealed objects.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.12 no.12
• /
• pp.5856-5864
• /
• 2011

In this paper, we propose a moving object detection method in road video sequences acquired from a stationary camera. Our proposed method is based on the background subtraction method using Gaussian pyramids in both the background images and input video frames. It is more effective than pixel based subtraction approaches to reduce false detections which come from the mis-registration between current frames and the background image. And to determine a threshold value automatically in subtracted images, we calculate the threshold value using Otsu's method in each frame and then apply a scalar Kalman filtering to the threshold value. Experimental results show that the proposed method effectively detects moving objects in road video images.