Color Saliency Map Research Articles

Display Line Defect Detection Method Based on Color Feature Fusion

Display color line defect detection is an important step in the production quality inspection process. In order to improve the detection accuracy of low contrast line defects, we propose a display line defect detection method based on color feature fusion. The color saliency channels in the RG|GR and BY|YB channels were obtained using the relative entropy maximum criterion. Then, RG|GR were combined with the a channel and BY|YB with the b channel to calculate the red-green and the blue-yellow color fusion maps. The fusion color saliency map of the red-green and the blue-yellow color fusion maps was obtained by color feature fusion. Finally, the segmentation threshold was calculated according to the mean and standard deviation of the fusion color saliency map. The fused color saliency map was binarized and segmented to obtain a binary map of color line defects. The experimental results show that for the detection of multi-background offline defects, the detection accuracy of the algorithm in this paper is better than 90%, while other mainstreams fail to detect. Compared with state-of-the-art saliency detection algorithms, our method is capable of real-time low-contrast line defect detection.

Region-based depth feature descriptor for saliency detection on light field

This paper addresses the light field saliency detection problem via a multiple cue integration framework. By reinterpreting the usage of dark channels in estimating the amount of defocus, a novel region-based depth feature descriptor (RDFD) defined over the focal stack is proposed. Compared to the methods which utilize the depth map as another image channel, the RDFD can produce more informative saliency cues and make less restrictive assumptions on accurate depth map or focused clear images containing dark pixels. The proposed RDFD facilitates saliency detection in the following two respects: (1) the region-based depth contrast map can be computed by measuring a pair-wise distance between super-pixels with the proposed RDFD, (2) a spatial distribution prior in the 3D space (3D-SDP) can be obtained from such depth measurements to provide high-level semantic guidances, including the gradient-like distribution in depth and the object-biased prior in image plane. Both of them contribute to generating a contrast-based depth saliency map and refining a background-based color saliency map. Finally, these saliency maps are merged into a single map using a multi-layer cellular antomata (MCA) optimizer. Experimental results demonstrate that our method outperforms state-of-the-art techniques on the challenging light field saliency detection benchmark LFSD.

Exploiting the Value of the Center-dark Channel Prior for Salient Object Detection

Saliency detection aims to detect the most attractive objects in images and is widely used as a foundation for various applications. In this article, we propose a novel salient object detection algorithm for RGB-D images using center-dark channel priors. First, we generate an initial saliency map based on a color saliency map and a depth saliency map of a given RGB-D image. Then, we generate a center-dark channel map based on center saliency and dark channel priors. Finally, we fuse the initial saliency map with the center dark channel map to generate the final saliency map. Extensive evaluations over four benchmark datasets demonstrate that our proposed method performs favorably against most of the state-of-the-art approaches. Besides, we further discuss the application of the proposed algorithm in small target detection and demonstrate the universal value of center-dark channel priors in the field of object detection.

Depth-aware saliency detection using convolutional neural networks

This paper proposes a new end-to-end depth-aware saliency model using three convolutional neural networks including color saliency network, depth saliency network and saliency fusion network, for saliency detection in RGBD images and stereoscopic images. Firstly, the color image is fed to the color saliency network to generate the color saliency map. Then, by sharing the weights of some layers in the color saliency network, the depth saliency network exploits the weight initialization and multi-layer pyramid feature fusion to learn effective depth features from the three-channel depth image, which is converted from the original depth map, and generates the depth saliency map. Finally, the saliency fusion network integrates the color saliency map with the depth saliency map into the final saliency map, which can highlight salient object regions and suppress background regions more effectively. Experimental results on five public datasets demonstrate that our model achieves the better performance compared with the state-of-the-art depth-aware saliency models.

Salient object detection employing regional principal color and texture cues

Saliency in a scene describes those facets of any stimulus that makes it stand out from the masses. Saliency detection has attracted numerous algorithms in recent past and proved to be an important aspect in object recognition, image compression, classification and retrieval tasks. The present method makes two complementary saliency maps namely color and texture. The method employs superpixel segmentation using Simple Linear Iterative Clustering (SLIC). The tiny regions obtained are further clustered on the basis of homogeneity using DBSCAN. The method also employs two levels of quantization of color that makes the saliency computation easier. Basically, it is an adaptation to the property of the human visual system by which it discards the less frequent colors in detecting the salient objects. Furthermore, color saliency map is computed using the center surround principle. For texture saliency map, Gabor filter is employed as it is proved to be one of the appropriate mechanisms for texture characterization. Finally, the color and texture saliency maps are combined in a non-linear manner to obtain the final saliency map. The experimental results along with the performance measures have established the efficacy of the proposed method.

Visual saliency guided complex image retrieval

ABSTRACT Compared with the traditional text data, multimedia data are concise and contains rich meanings, so people are more willing to use the multimedia data to store information. How to effectively retrieve information is essential. This paper proposes a novel visual saliency guided complex image retrieval model. Initially, Itti visual saliency model is presented. In this model, the overall saliency map is generated by the integration of direction, intensity and color saliency map, respectively. Then, to help describe the image pattern more clearly, we present the multi-feature fusion paradigm of images. To address the complexity of the images, we propose a two-stage definition: (1) Cognitive load based complexity; (2) Cognitive level of complexity classification. The group sparse logistic regression model is integrated to finalize the image retrieval system. The performance of the proposed system is tested on different databases compared with the other state-of-the-art models which overcome the baselines in complex scenarios.

Saliency detection method fused depth information based on Bayesian framework

In the complex background, the traditional saliency detection methods often encounter the problems of unstable detection results and low accuracy. To address this problem, a saliency detection method fused depth information based on Bayesian framework is proposed. Firstly, the color saliency map is obtained by using a variety of contrast methods which includes global contrast, local contrast and foreground-background contrast, and the depth saliency map is obtained by using the depth contrast method based on the anisotropic center-surround difference. Secondly, using the Bayesian model to fuse the color-based saliency map and the depth-based saliency map. The experimental results show that the proposed method can effectively detect the salient targets under complex background and achieve higher detection accuracy on the published NLPR-RGBD dataset and NJU-DS400 dataset.

Automatic segmentation of dermoscopy images using saliency combined with Otsu threshold

Segmentation is one of the crucial steps for the computer-aided diagnosis (CAD) of skin cancer with dermoscopy images. To accurately extract lesion borders from dermoscopy images, a novel automatic segmentation algorithm using saliency combined with Otsu threshold is proposed in this paper, which includes enhancement and segmentation stages. In the enhancement stage, prior information on healthy skin is extracted, and the color saliency map and brightness saliency map are constructed respectively. By fusing the two saliency maps, the final enhanced image is obtained. In the segmentation stage, according to the histogram distribution of the enhanced image, an optimization function is designed to adjust the traditional Otsu threshold method to obtain more accurate lesion borders. The proposed model is validated from enhancement effectiveness and segmentation accuracy. Experimental results demonstrate that our method is robust and performs better than other state-of-the-art methods.

Saliency Detection for Small Maritime Target Using Singular Value Decomposition of Amplitude Spectrum

ABSTRACTSpectrum analysis based saliency detection approaches have been studied and applied in computer vision. They are either generalized filtering or spectrum whitening of amplitude spectrum of image. However, as resultant salient regions may be inconsistent, a better result can be achieved by combining several scales. This paper proposes a novel saliency detection method for small maritime target using singular value decomposition (SVD) of amplitude spectrum. Color and intensity features of image are extracted and Fourier transform is employed on them. SVD is conducted on amplitude spectrums of each feature and their saliency maps are defined as the inverse Fourier transform of non-dominant components, whose amplitude spectrum is measured by difference between original amplitude spectrums and their main components. Taking into account the characteristics of maritime visual scene, the dominant parts of amplitude spectrum are selected according to threshold which is the ratio of singular values. Color saliency map and intensity saliency map are linearly combined to achieve the saliency map. The proposed method is conducted on the original image size, thus maintaining information of small maritime target. Experiments are made to show the validity and effectiveness. Comparison between proposed method and related approaches demonstrates that the new method is more suitable for saliency detection of small maritime target. Effect of threshold is also analyzed, and 50% is best for saliency detection of small maritime targets. The new saliency detection approach can be used for small maritime target detection and tracking.

Unified Saliency Detection Model Using Color and Texture Features.

Saliency detection attracted attention of many researchers and had become a very active area of research. Recently, many saliency detection models have been proposed and achieved excellent performance in various fields. However, most of these models only consider low-level features. This paper proposes a novel saliency detection model using both color and texture features and incorporating higher-level priors. The SLIC superpixel algorithm is applied to form an over-segmentation of the image. Color saliency map and texture saliency map are calculated based on the region contrast method and adaptive weight. Higher-level priors including location prior and color prior are incorporated into the model to achieve a better performance and full resolution saliency map is obtained by using the up-sampling method. Experimental results on three datasets demonstrate that the proposed saliency detection model outperforms the state-of-the-art models.

Interactive Segmentation of Color Images Based on Color Saliency

SUMMARYEfficient interactive foreground extraction from color images is of great practical importance in computer vision. In recent years, an approach based on optimization using graph‐cuts has been widely used. However, an interactive foreground extraction approach, which is intuitive and efficient for users, is required for practical applications. In this paper, we propose a novel interactive foreground extraction method based on color saliency. In our method, the user provides only some foreground pixels as the initial reference. To achieve extraction of multiple target objects, the likelihoods of the foreground and background are defined by a Gaussian mixture model and a color saliency map based on the provided reference pixels. These likelihoods are added to the cost of the proposed graph. Finally, image segmentation is performed by optimizing the proposed graph using the graph‐cut algorithm.

Structure Lane Detection Based on Saliency Feature of Color and Direction

In this paper, we propose a method for structure lane detection; the method is based on two features: color and direction. This method can improve the robustness and accuracy of lane detection. Two kinds of saliency map have been calculated: color saliency map and direction saliency map. The final saliency map is the combination of the two map mentioned above. The binary image is getting from the final saliency map, and the feature points which used for fitting have been selected. The road region is segmented by the lanes. Experiment result shows that the proposed method produces better performance against some other methods.

Foreground Object Extraction Based on Interactive Color Saliency Map

Efficient interactive foreground extraction from color images is of great practical importance in computer vision. In recent years, an approach based on optimization using graph-cuts has been widely used. However, an interactive foreground extraction approach, which is intuitive and efficient for users, is required for practical applications. In this paper, we propose a novel foreground object extraction method based on interactive color saliency map. In our method, the user provides only some foreground pixels as the initial reference. To achieve extraction of multi target objects, likelihoods of the foreground and background are defined by a color saliency map based on the provided reference pixels. These likelihoods are added to the cost of the proposed graph. Finally, image segmentation is performed by optimizing the proposed graph using the graph-cut algorithm.

Interactive Segmentation for Color Image based on Color Saliency

Efficient interactive foreground extraction from color images is of great practical importance in computer vision. In recent yaers, an approach, based on optimization using graph-cuts has been widely used. However, an interactive foreground extraction approach, which is intuitive and efficient for users, is required for practical applications. In this paper, we propose a novel interactive foreground extraction method based on color saliency. In our method, the user provides only some foreground pixels as the initial reference. To achieve extraction of multi target objects, likelihoods of the foreground and background are defined by a gaussian mixture model and a color saliency map based on the provided reference pixels. These likelihoods are added to the cost of the proposed graph. Finally, image segmentation is performed by optimizing the proposed graph using the graph-cut algorithm.

Biologically Inspired Particle Filter for Robust Visual Tracking1

Although particle filter and its variants like KPF and UPF achieve great success in many visual tracking applications, they depend on local proposal distributions and hence always fall flat in cases of global object shift and large-scale object movement. To address this issue, we present the concept of global proposal distribution for particle filter with the inspiration from biological vision examples and term their combination as biologically inspired particle filter (BI-PF). Three types of global proposal distributions are proposed based on biological vision inspired saliency maps, including motion, color, and contrast saliency map. In order for BI-PF to mimic the way human’s eye tracks object, local and global proposal distribution are dynamically integrated through a threshold determining when to switch between them. Experimental results demonstrate the outperformance of our BI-PFs over particle filter, KPF, and UPF. Keywards: Visual Tracking, Particle Filter, Proposal Distribution, Biological Vision, Saliency Map

Face Image Segmentation Using Color Information and Saliency Map

In this paper, we proposed an efficient method to address the problem of color face image segmentation that is based on color information and saliency map. This method consists of three stages. At first, skin colored regions is detected using a Bayesian model of the human skin color. Then, we get a chroma chart that shows likelihoods of skin colors. This chroma chart is further segmented into skin region that satisfy the homogeneity property of the human skin. The third stage, visual attention model are employed to localize the face region according to the saliency map while the bottom-up approach utilizes both the intensity and color features maps from the test image. Experimental evaluation on test shows that the proposed method is capable of segmenting the face area quite effectively，at the same time, our methods shows good performance for subjects in both simple and complex backgrounds, as well as varying illumination conditions and skin color variances.

Salient Region Detection by Modeling Distributions of Color and Orientation

We present a robust salient region detection framework based on the color and orientation distribution in images. The proposed framework consists of a color saliency framework and an orientation saliency framework. The color saliency framework detects salient regions based on the spatial distribution of the component colors in the image space and their remoteness in the color space. The dominant hues in the image are used to initialize an expectation-maximization (EM) algorithm to fit a Gaussian mixture model in the hue-saturation (H-S) space. The mixture of Gaussians framework in H-S space is used to compute the inter-cluster distance in the H-S domain as well as the relative spread among the corresponding colors in the spatial domain. Orientation saliency framework detects salient regions in images based on the global and local behavior of different orientations in the image. The oriented spectral information from the Fourier transform of the local patches in the image is used to obtain the local orientation histogram of the image. Salient regions are further detected by identifying spatially confined orientations and with the local patches that possess high orientation entropy contrast. The final saliency map is selected as either color saliency map or orientation saliency map by automatically identifying which of the maps leads to the correct identification of the salient region. The experiments are carried out on a large image database annotated with ldquoground-truthrdquo salient regions, provided by Microsoft Research Asia, which enables us to conduct robust objective level comparisons with other salient region detection algorithms.