Robust image alignment is a necessary and challenging step for numerous computational photography applications. In particular, large camera motion poses significant challenge to Mobile High Dynamic Range (HDR) Imaging...
详细信息
ISBN:
(纸本)9781509053162
Robust image alignment is a necessary and challenging step for numerous computational photography applications. In particular, large camera motion poses significant challenge to Mobile High Dynamic Range (HDR) Imaging due to hand-held capture of input images and limited computational resources. Aligning images only by detecting and matching image features is computationally expensive and can also be erratic. We present a robust multi-sensory method for aligning exposure bracketed images on mobile cameras. We use inertial sensor based camera pose estimate to pre-warp images and iteratively align them by minimizing alignment error. We also simultaneously estimate local motion masks which can be used to eliminate ghosting artifacts in the final HDR image. We collected HDR image dataset with diverse scenes along with inertial sensor data, which is a novel contribution and have evaluated our performance with existing mobile HDR image alignment techniques in literature.
This paper is concerned with experimental analysis of visual inverted pendulum servoing system. Firstly, visual inverted pendulum servoing system is introduced, and three typical imageprocessing algorithms are descri...
详细信息
ISBN:
(纸本)9789811026690;9789811026683
This paper is concerned with experimental analysis of visual inverted pendulum servoing system. Firstly, visual inverted pendulum servoing system is introduced, and three typical imageprocessing algorithms are described. These three algorithms are then employed to process the image of inverted pendulum captured by camera. Comparative experiments are operated, and the detection precision and real time performance are analyzed. This lays a solid foundation for future control research of visual inverted pendulum servoing system.
Curvature, the second-order directional derivative of an image, has been widely used for image interpolation. However, conventional curvature-based interpolation (CBI) methods employ a time-consuming post processing t...
详细信息
ISBN:
(纸本)9781467383646
Curvature, the second-order directional derivative of an image, has been widely used for image interpolation. However, conventional curvature-based interpolation (CBI) methods employ a time-consuming post processing to reduce visual artifacts such as blurring and jagging caused by the inaccurate estimation of the local curvature. In this paper, a novel fast CBI method is proposed which can precisely estimate the local curvature using the geometric relationship between the curvatures of the low-resolution and high-resolution images. Experimental results show that the proposed method is not only robust to visual artifacts even without any post processing, but also about 40 times faster than the conventional CBI method, making it applicable to consumer electronics.
An improved model for additive spread spectrum (SS) watermark detection on compressed sensing (CS) reconstructed images is presented in this paper. Mathematical form of detection threshold in log-likelihood ratio mode...
详细信息
ISBN:
(纸本)9781509017461
An improved model for additive spread spectrum (SS) watermark detection on compressed sensing (CS) reconstructed images is presented in this paper. Mathematical form of detection threshold in log-likelihood ratio model is derived first and it is seen that detection probability depends on the embedding strength, watermark power, host signal variance on CS along with noise variance in observation/measurements. An optimization framework is then developed to minimize the visual distortion that includes reconstruction and embedding distortion while satisfying certain detection reliability constraint. An approximate closed form solution to the optimization problem in terms of embedding strength and set of appropriate host samples selection for a given number of CS measurements is derived and validated by a large set of simulations.
Existing no-reference image quality assessment (NR-IQA) methods mainly focus on designing the low-level features related to image degradation. However, the evaluation of image quality is the human visual perception of...
详细信息
ISBN:
(纸本)9781509053162
Existing no-reference image quality assessment (NR-IQA) methods mainly focus on designing the low-level features related to image degradation. However, the evaluation of image quality is the human visual perception of image content, involving the integrated analysis of global high-level semantics and local low-level characteristics. From this perspective, we propose a NR-IQA framework based on global and local content perception. We adopt the deep convolutional neural network (DCNN) to extract the semantic feature implied in global image content. The perception of visual quality associated with local content utilizes the visual attention and filtering mechanisms of human visual system. The overall image quality is estimated by combining the semantic and local characteristic features generated from the perceptions. Experimental results on the LIVE IQA database demonstrate that our method is superior to the state-of-the-art NR-IQA algorithms and competitive to the popular full reference IQA methods. Further experiments on the TID2008 dataset show that the proposed approach is robust for various kinds of distortion types.
Object selection is a challenge in computer vision since it is generally a trade-off between accuracy and performance. A popular approach is the use of bounding boxes around objects that are to be selected. Other comm...
详细信息
ISBN:
(纸本)9781509053162
Object selection is a challenge in computer vision since it is generally a trade-off between accuracy and performance. A popular approach is the use of bounding boxes around objects that are to be selected. Other common techniques provide a set of objects from which the user can then choose. The method presented in this paper is designed around the priority of performance and granular selection of objects in the scene. Experiments performed on a non-parallel implementation of the proposed solution return results in an average time of 0.043s. The technique also returned very good results in the processing of objects that are partially occluded, hence enabling future work in improved identification and recognition of such objects.
This work explores the subject of thermal imagery in the context of face recognition. Its aim is to create a database of facial images taken in both thermal and visual domains. To achieve this, a specialized photograp...
详细信息
ISBN:
(纸本)9783319238142;9783319238135
This work explores the subject of thermal imagery in the context of face recognition. Its aim is to create a database of facial images taken in both thermal and visual domains. To achieve this, a specialized photographic stand was designed, which allows simultaneous capture of images from IR thermal camera and SLR digital camera. To ensure precision, stability and fluency of photographic sessions, a Matlab application has been developed, through which it is possible to remotely control both devices, as well as automatically download captured images onto a hard drive and save them within an organized folder structure. Additionally, several image fusion techniques have been implemented in order to effectively combine visual and thermal data for use in face recognition algorithms.
Single image haze removal is an underdetermined inverse problem whose solution hinges on valid image priors or models. In this work, robust priors drawn from outdoor scene statistics are explored. Specifically, a Gaus...
详细信息
ISBN:
(纸本)9781509053162
Single image haze removal is an underdetermined inverse problem whose solution hinges on valid image priors or models. In this work, robust priors drawn from outdoor scene statistics are explored. Specifically, a Gaussian mixture model of chrominance distribution is proposed toward transmittance estimation and its physical validity is justified. In addition, a new sparsity-based optimization approach for transmittance image super-resolution/restoration is proposed, which makes a solid assumption that most outdoor object surfaces are piece wise linear and thus the corresponding depth image is sparse in Laplacian space. Experimental results are given in proof of the remarkably improved visual quality of our new haze removal technique over its predecessors.
Nowadays visual attention has been applied to many research and application problems. Different algorithms from low level to high level have been developed to detect the saliency map. For images with human face, high-...
详细信息
ISBN:
(纸本)9781509053162
Nowadays visual attention has been applied to many research and application problems. Different algorithms from low level to high level have been developed to detect the saliency map. For images with human face, high-level factors like mole may influence the visual attention. To investigate visual attention on human face with mole, we construct a visual Attention database for Faces with Mole (VAFM) that contains face images, fixation density maps (FDM), landmark points as well as eye tracking data. Then we build visual attention model for face images with mole combining low-level saliency algorithms and high-level feature. Compared with the traditional low-level saliency algorithms, the proposed model perform better on our dataset.
Saliency analysis is an important implement for remote sensing imageprocessing. It can effectively solve the contradiction between accuracy and computation complexity when it is applied in the region of interest (ROI...
详细信息
ISBN:
(纸本)9781509053162
Saliency analysis is an important implement for remote sensing imageprocessing. It can effectively solve the contradiction between accuracy and computation complexity when it is applied in the region of interest (ROI) detection and extraction for remote sensing images. In this paper, we propose an efficient ROI detection model for remote sensing images based on low-level contrast feature saliency analysis. For the proposed model, we first perform fast directional integer wavelet transform (FD-IWT) to obtain multi-scale approximate and detail coefficients. Then these multi-scale orientation, local contrast, and global contrast features are exploited to generate the saliency map. Qualitative and quantitative evaluation shows that the proposed model outperforms the other nine state-of-art ROI detection models for that the proposed model can obtain highlighted and integrated ROI with well-defined boundaries, as well as eliminate the shadow interference in remote sensing image.
暂无评论