Variation in pose is one of the main obstacles confronting researchers in the area of face recognition. In this paper, a novel method is proposed to explicitly tackle this problem. Multi-color uniform local binary pat...
详细信息
Variation in pose is one of the main obstacles confronting researchers in the area of face recognition. In this paper, a novel method is proposed to explicitly tackle this problem. Multi-color uniform local binary pattern (ULBP) is introduced for extracting salient features along with wavelet transform. Learning scheme is adopted to obtain a mapping coefficient vector between face in a pose and frontal face. Then expected frontal face view vector could be generated by inserting the posed face. Instead of using the entire face, some of its important regions are taken into account. The proposed method relies only on single frontal face image as a gallery image. Results have demonstrated that the proposed method operates well even under the low-resolution conditions.
作者:
Seyed Omid ShahdiS.A.R. Abu-BakarComputer Vision
Video and Image Processing (CvviP) Research Laboratory Faculty of Electrical Engineering Universiti Teknologi Malaysia Skudai Johor Malaysia
Pose variations are known to give real challenges in face recognition system. In this paper we proposed a method to recognize non-frontal faces with high performance by relying only on single full frontal gallery face...
详细信息
Pose variations are known to give real challenges in face recognition system. In this paper we proposed a method to recognize non-frontal faces with high performance by relying only on single full frontal gallery faces. By utilizing only small regions of the face or patches, we compute the Fourier coefficients of these patches for each image and transform them into a single vector. Hence, instead of comparing and matching pixels values we use these vectors to form a linear relationship which is then used to estimate the frontal face vector and then compare it with the actual frontal feature vector. The results show an average performance accuracy of 90% across all pose.
This paper examines the effectiveness of geometric feature descriptors, common in computervision, for false positive reduction and for classification of lung nodules in low dose CT (LDCT) scans. A data-driven lung no...
详细信息
This paper examines the effectiveness of geometric feature descriptors, common in computervision, for false positive reduction and for classification of lung nodules in low dose CT (LDCT) scans. A data-driven lung nodule modeling approach creates templates for common nodule types, using active appearance models (AAM); which are then used to detect candidate nodules based on optimum similarity measured by the normalized cross-correlation (NCC). Geometric feature descriptors (e.g., SIFT, LBP and SURF) are applied to the output of the detection step, in order to extract features from the nodule candidates, for further enhancement of output and possible reduction of false positives. Results on the clinical ELCAP database showed that the descriptors provide 2% enhancements in the specificity of the detected nodule above the NCC results when used in a k-NN classifier. Thus quantitative measures of enhancements of the performance of CAD models based on LDCT are now possible and are entirely model-based. Most importantly, our approach is applicable for classification of nodules into categories and pathologies.
This paper deals with the problem of estimating structure of 3D scenes and image transformations from observations that are blurred due to unconstrained camera motion. Initially, we consider a fronto-parallel planar s...
详细信息
Unrestricted camera motion and the ability to operate over a range of lens parameters are often desirable when using an off-the-shelf camera. Variations in intrinsic and extrinsic parameters induce defocus and pixel m...
详细信息
In images and videos of a 3D scene, blur due to camera shake can be a source of depth information. Our objective is to find the shape of the scene from its motion-blurred observations without having to restore the ori...
详细信息
作者:
S. SetuminU.U. SheikhS.A.R Abu-BakarComputer Vision
Video and Image Processing Lab Department of Microelectronics and Computer Engineering Faculty of Electrical Engineering Universiti Teknologi Malaysia Malaysia
In this paper we address the issue of recognizing nonstandard Malaysian car license plates. These plates contain nonstandard characters such as italic, cursive and connected letters, which most plate recognition syste...
详细信息
In this paper we address the issue of recognizing nonstandard Malaysian car license plates. These plates contain nonstandard characters such as italic, cursive and connected letters, which most plate recognition systems are unable to recognize. We propose a technique using stroke extraction and analysis to recognize these nonstandard characters. The proposed technique first extracts the contour of the character and then the stroke direction which are used for classification. The advantage of this method is that the system requires no training. From the experiments performed, the method has a correct recognition accuracy of 95% and even works with standard car plates.
The quality inspection for export star fruit is still perform manually by human labor until today. Due to manual process, a real-time system for star fruit color maturity inspection is developed in this paper. In real...
详细信息
The quality inspection for export star fruit is still perform manually by human labor until today. Due to manual process, a real-time system for star fruit color maturity inspection is developed in this paper. In real-time application, most of the image acquisition device is using YCbCr color space data such as CCD camera. This paper presents the modification on the previous star fruit color maturity classification algorithm which is based on RGB color space into YCbCr color space. In this new modified algorithm, the system is faster than before because the star fruit maturity classification process operates without any mathematical operation involved in the feature extraction process. This process is possible as the color information can be obtained directly from the Cb and Cr component which is not the case in the RGB color space. Based on the experiment results, the classification accuracy for the modified algorithm is 96%.
Segmentation in a classification system plays an important role as it will determine the region of interest for the classification. In this work, the segmentation of star fruit is performed in YCbCr color space. Becau...
详细信息
Segmentation in a classification system plays an important role as it will determine the region of interest for the classification. In this work, the segmentation of star fruit is performed in YCbCr color space. Because of the background of the image to be classify has 2 different colors, which are black and white, the segmentation is obtained using a fixed threshold value for the Cb component. This is because the value for Cb is equivalent in black, white or gray scale pixels. By fixing the threshold for Cb component, the region of interest (ROI) can easily be determined and then proceeded to the classification process. The results in this paper show that the ROI can be determined completely by this method.
作者:
S. SetuminU.U. SheikhS.A.R Abu-BakarComputer Vision
Video and Image Processing Research Lab Department of Microelectronics and Computer Engineering Faculty of Electrical Engineering Universiti Teknologi Malaysia Malaysia
In this paper we address the issue of locating non-standard Malaysian car license plate. Instead of searching the region for the plate, we directly locate the alphanumeric characters of the car plate. In this manner, ...
In this paper we address the issue of locating non-standard Malaysian car license plate. Instead of searching the region for the plate, we directly locate the alphanumeric characters of the car plate. In this manner, we remove issues such as plate size variations and plates on black colored vehicles. Our main goal is to locate and extract the alphanumeric characters of Malaysian special plates. These special plates do not follow the normal standard car plates' format as they may contain italic, cursive and connected letters, and of different fonts. Using several parameters such as pixel compactness, angles, and projection histogram, we use ruled-based technique to locate and detect these special characters of the car plates. The results have shown that we are able to automatically locate with an accuracy of 95%.
暂无评论