Deng-Yuan Huang

Moving object detection and tracking from video captured by moving camera

Proposed method has good performance for a moving camera without additional sensors.Proposed method works well for tracking overlapping objects with scale changes.Proposed method outperforms the-state-of-art methods. This paper presents an effective method for the detection and tracking of multiple moving objects from a video sequence captured by a moving camera without additional sensors. Moving object detection is relatively difficult for video captured by a moving camera, since camera motion and object motion are mixed. In the proposed method, the feature points in the frames are found and then classified as belonging to foreground or background features. Next, moving object regions are obtained using an integration scheme based on foreground feature points and foreground regions, which are obtained using an image difference scheme. Then, a compensation scheme based on the motion history of the continuous motion contours obtained from three consecutive frames is applied to increase the regions of moving objects. Moving objects are detected using a refinement scheme and a minimum bounding box. Finally, moving object tracking is achieved using a Kalman filter based on the center of gravity of a moving object region in the minimum bounding box. Experimental results show that the proposed method has good performance.

Robust real-time ship detection and tracking for visual surveillance of cage aquaculture

Abstract This paper presents a visual surveillance scheme for cage aquaculture that automatically detects and tracks ships (intruders). For ship detection and tracking, we propose a robust foreground detection and background updating to effectively reduce the influence of sea waves. Furthermore, we propose a fast 4-connected component labeling method to greatly reduce the computational cost associated with the conventional method. Wave ripples are removed from regions with ships. An improved full search algorithm based on adaptive template block matching with a wave ripple removal is presented to quickly, accurately, and reliably track overlapping ships whose scales change. Experimental results demonstrate that the proposed schemes have outstanding performance in ship detection and tracking. The proposed visual surveillance system for cage aquaculture triggers an alarm if intruders are detected. The security of cage aquaculture can be increased. The proposed visual surveillance can thus greatly help the popularization of cage aquaculture for ocean farming.

Reliable moving vehicle detection based on the filtering of swinging tree leaves and raindrops

An efficient method for detecting moving vehicles based on the filtering of swinging trees and raindrops is proposed. To extract moving objects from the background, an adaptive background subtraction scheme with a shadow elimination model is used. Swinging trees are removed from foreground objects to reduce the computational complexity of subsequent tracking. Raindrops are removed from foreground objects when necessary. Performance evaluations are carried out using seven real-world traffic image sequences. Experimental results show average recognition rates of 96.83% and 97.20% for swinging trees and raindrops, respectively, indicating the feasibility of the proposed method.

Effective image forgery detection of tampered foreground or background image based on image watermarking and alpha mattes

This paper proposes an effective image forgery detection scheme that identifies a tampered foreground or background image using image watermarking and alpha mattes. In the proposed method, component-hue-difference-based spectral matting is used to obtain the foreground and background images based on the obtained alpha matte. Next, image watermarking based on the discrete wavelet transform, discrete cosine transform, and singular value decomposition is used to embed two different watermarks into the foreground and background images, respectively. Finally, the difference between the obtained singular values is used to detect tampering of foreground or background image. Experimental results show that the proposed method performs well in terms of image forgery detection.

Rapid detection of camera tampering and abnormal disturbance for video surveillance system

Camera tampering and abnormalities are examined for video surveillance system.Brightness, edge details, and histogram information are computationally efficient.The system runs at 20-30frames/s, meeting the requirement of real-time operation.An average of 4.4% of missed events indicates the feasibility of proposed method. Camera tampering may indicate that a criminal act is occurring. Common examples of camera tampering are turning the camera lens to point to a different direction (i.e., camera motion) and covering the lens by opaque objects or with paint (i.e., camera occlusion). Moreover, various abnormalities such as screen shaking, fogging, defocus, color cast, and screen flickering can strongly deteriorate the performance of a video surveillance system. This study proposes an automated method for rapidly detecting camera tampering and various abnormalities for a video surveillance system. The proposed method is based on the analyses of brightness, edge details, histogram distribution, and high-frequency information, making it computationally efficient. The proposed system runs at a frame rate of 20-30frames/s, meeting the requirement of real-time operation. Experimental results show the superiority of the proposed method with an average of 4.4% of missed events compared to existing works.

Video object segmentation in rainy situations based on difference scheme with object structure and color analysis

A scheme based on a difference scheme using object structures and color analysis is proposed for video object segmentation in rainy situations. Since shadows and color reflections on the wet ground pose problems for conventional video object segmentation, the proposed method combines the background construction-based video object segmentation and the foreground extraction-based video object segmentation where pixels in both the foreground and background from a video sequence are separated using histogram-based change detection from which the background can be constructed and detection of the initial moving object masks based on a frame difference mask and a background subtraction mask can be further used to obtain coarse object regions. Shadow regions and color-reflection regions on the wet ground are removed from the initial moving object masks via a diamond window mask and color analysis of the moving object. Finally, the boundary of the moving object is refined using connected component labeling and morphological operations. Experimental results show that the proposed method performs well for video object segmentation in rainy situations.

Gait recognition based on Gabor wavelets and modified gait energy image for human identification

Abstract. This paper proposes a method for recognizing human identity using gait features based on Gabor wavelets and modified gait energy images (GEIs). Identity recognition by gait generally involves gait representation, extraction, and classification. In this work, a modified GEI convolved with an ensemble of Gabor wavelets is proposed as a gait feature. Principal component analysis is then used to project the Gabor-wavelet-based gait features into a lower-dimension feature space for subsequent classification. Finally, support vector machine classifiers based on a radial basis function kernel are trained and utilized to recognize human identity. The major contributions of this paper are as follows: (1) the consideration of the shadow effect to yield a more complete segmentation of gait silhouettes; (2) the utilization of motion estimation to track people when walkers overlap; and (3) the derivation of modified GEIs to extract more useful gait information. Extensive performance evaluation shows a great improvement of recognition accuracy due to the use of shadow removal, motion estimation, and gait representation using the modified GEIs and Gabor wavelets.

Face recognition based on collaborative image similarity assessment

In this paper, we propose an innovative approach for face recognition based on collaborative image similarity assessment (CISA). In the proposed method, the test sample is first represented by a linear combination of all the training samples for each face class. The classification task is then conducted using the similarity measures including structure similarity index measure (SSIM), root mean square (RMS), and similarity assessment value (SAV). Since CISA is only one phase, it is computationally efficient when comparing with the method of two-phase test sample sparse representation (TPTSR). To verify the performance of face classification, two popular face databases of the ORL and FERET are used. Results show that CISA is comparable with TPTSR on the classification rates for ORL database. However, CISA greatly outperforms TPTSR on the evaluation of the FERET database. Moreover, only 276.4ms on an average is required for CISA in the classification of each test sample but it needs 800.8ms for TPTSR.

Boosting Scheme for Detecting Region Duplication Forgery in Digital Images

The detection of copy-move forgery image is important in the field of blind image forensics because it is of pure image processing technique without any support of embedded security information. The proposed method consists of a boosting scheme, feature extraction and similarity matching for the detection of duplicated regions. The boosting scheme comprises an estimation of dark channel, histogram equalization and grayscale layering, by which the number of image blocks on each subimage layer can be dramatically reduced so that the time efficiency of subsequent lexicographical sorting and similarity matching can be greatly improved. Experimental results show that the proposed boosting scheme can significantly enhance the computation efficiency and have a good detection rate. Moreover, the propose method is robust to any angles rotation attack.

Eye Detection Based on Skin Color Analysis with Different Poses under Varying Illumination Environment

In this paper, we propose a method of eye detection based on skin color analysis under varying illumination. The proposed method consists of several phases, including color conversion, skin color segmentation and face mask calculation, facial feature extraction and eye candidate determination, and detection of human eyes. To eliminate the effect of lighting change on the performance of eye detection, color conversion is first performed. Face mask calculation based on skin color segmentation is then carried out to reduce the possible searching region of human eyes. Eye candidates detected using facial features are used to define the possible human eyes. Human eyes are thus detected by the geometric features of gravity and spatial centers for these eye candidates. Results show that the proposed method works well for faces with different poses and multiple faces under varying illumination. The eye detection time of 21.8 ms is achieved for an image of size 213x320 pixels, indicating computational efficiency of the proposed system.