An improved feature extraction-based colour image matching method using quaternion matrix

High-speed color digital imaging devices are desired for many applications, including civilian and military missions. To capture color images a color sensor has to divide the incidence light into three spectral bands: blue, green and red. Due to the limitation of the spectral bandwidth, the speed of a color digital sensor is significant slower than that of a black and white sensor covering the entire visible bandwidth or the entire bandwidth from visible to near infrared. This paper introduces the conception of a high-speed color imaging device which can significantly increase the imaging speed without losing image resolution and other image qualities. This device uses the concept of black and white imaging to increase the imaging speed while keeping the image resolution unchanged. It also uses the concept of color imaging to capture color information, but at a significantly lower resolution to allow more incidence light being received by each sensor detector (pixel) to increase the imaging speed. The final color image at the desired resolution is recovered by integrating the spatial information from the black and white image and the color information from the low-resolution color image. An automatic image fusion technique is used to integrate the spatial information and the color information. The quality of the image fusion technique plays a crucial role for quality of the recovered color image. The concept of this high-speed color imaging device has been tested with available low-resolution color images and high-resolution black and white images from the latest commercial satellites, Ikonos and QuickBird. The test results demonstrated that the concept of the high-speed digital color imaging device is promising for producing real high-speed color digital devices for practical applications.

In this paper, we study color image inpainting as a pure quaternion matrix completion problem. In the literature, the theoretical guarantee for quaternion matrix completion is not well-established. Our main aim is to propose a new minimization problem with an objective combining nuclear norm and a quadratic loss weighted among three channels. To fill the theoretical vacancy, we obtain the error bound in both clean and corrupted regimes, which relies on some new results of quaternion matrices. A general Gaussian noise is considered in robust completion where all observations are corrupted. Motivated by the error bound, we propose to handle unbalanced or correlated noise via a cross-channel weight in the quadratic loss, with the main purpose of rebalancing noise level, or removing noise correlation. Extensive experimental results on synthetic and color image data are presented to confirm and demonstrate our theoretical findings.

The paper proposes a novel method for estimating the optical flows from the sequentially captured images with using their own color information. The gradient method is well known as one of the conventional methods to estimate the flows, and then the spatial and temporal derivative of the images are used in the method. Since the color images have richer information than the monochrome ones, they should contribute for estimating the more precise optical flows. In our approach, the color derivative vector (CDV) is introduced to bring out the information in the color images for the optical flow estimation. The CDV is derived from the derivatives of color images, and the optimal CDV provides the concrete weighting values of the RGB data. The optimal CDV is obtained with using the eigenvalues and the eigenvectors of the matrix consisting of the spatial and temporal color derivatives.

In this paper, local variance is used to describe the structural information of a color image in order to assess its quality. The representation method is different from conventional models in that some information that is sensitive to human eyes is enhanced by using local variance distribution. It encodes the local variance distribution of different channels of a color image into the three imaginary parts of a quaternion. The distance between the singular value feature vectors of the source image block and the distorted image block which are described by quaternion matrices is calculated. The experimental results show that the assessment results of the proposed assessment method are more consistent with the Human Visual System than those of the conventional assessment methods.

Most change detection methods are based on gray-level images. A gray-level image is regarded as a 1-D projection of three channels of color images. Therefore, more precise change detection results are expected by utilizing color information. We previously developed a change detection scheme using color images. In this paper, we determine which color space should be selected for accurate change detection based on our previous detection scheme. Our method can be applied to various color spaces, including gray-level images. Then we can measure the expected number of error pixels in order to select an appropriate color space which gives the best result among various color spaces. The experiments show that selecting a color space based on measurements results in the fewest error pixels.

Color and Depth-Based Superpixels for Background and Object Segmentation

Purpose: To automatically and efficiently count cell‐clusters of a clonogenic assay using color information and image processing techniques. Method and Materials: A computer algorithm was designed in the Matlab (Natick, MA) programming environment to automatically count cell colonies. In this implementation, six‐well tissue culture plates with stained colonies were scanned using a commercial flatbed scanner. These images were then read into the program as three‐channel (Red, Green, Blue) color images. Color windows were applied to each channel to isolate colony clusters from the background. Region labeling techniques were then used to identify individual cell clusters. Next, individual cluster statistics (area, eccentricity, etc.) were obtained and used to determine if a cluster were a viable colony (>50 cells) and whether it was a single, double or triple colony. The number of colonies counted was compared with the number determined manually. Results: The images obtained from the flatbed scanner were of high quality and were observed to have minimal shadowing effects. The time for the automated colony counting software to analyze and count individual wells was on the order of 7–20 sec. Comparison between manual and computer‐assisted counting showed agreement of 3–5%. The Wilcoxon signed‐rank test showed no statistically significant difference (p > 0.05) between the number of colonies counted manually or with the computer program. Conclusion:Images from the flatbed scanner effectively represent the required colony data, and no advanced equipment is required to record the images. Color range filtering removed the need for most post‐processing techniques and also is an effective method for dealing with errors from shadowing along the edges of the plates. Use of this automated technique will allow for a large quantity of clonogenic survival data to be analyzed in an efficient and consistent manner.

Because of the advantages of deep learning and information fusion technology, it has drawn much attention for researchers to combine them to achieve target recognition, positioning, and tracking. However, when the existing neural network process multichannel images (e.g., color images), multiple channels as a whole input into neural networks, which makes it hard for networks to fully learn information in R, G, and B channels of images. Therefore, it is not conducive to the final learning effect of the networks. To solve the problem, using different combinations of R, G, and B channels of color images for feature-level fusion, this paper proposes three fusion types as “R/G/B”, “R+G/G+B/B+R”, and “R+G+B/R+G+B/R+G+B” multichannel concat-fusional convolutional neural networks. Experimental results show that multichannel concat-fusional convolutional neural networks with fusional types of “R+G/G+B/B+R” and “R+G+B/R+G+B/R+G+B” achieve better performance than the corresponding non-fusional convolutional neural networks on different datasets. It shows that networks with fusion types of “R+G/G+B/B+R” and “R+G+B/R+G+B/R+G+B” can learn more fully information of R, G, and B channels of color images and improve the learning performance of networks.

The reliability and security of multimedia contents in transmission, communications, storage, and usage have attracted special attention. Robust image hashing, also referred to as perceptual image hashing, is widely applied in multimedia authentication and forensics, image retrieval, image indexing, and digital image watermarking. In this work, a novel robust image hashing method based on quaternion Zernike moments (QZMs) is proposed. QZMs offer a sound way to jointly deal with the three channels of color images without discarding chrominance information; the generated hash is thus shorter than the hash of three channels separately processing. The proposed approach's performance was evaluated on the color images database of UCID and compared with several recent and efficient methods. These experiments show that the proposed scheme provides a short hash in length that is robust to most common image content-preserving manipulations like JPEG compression, filtering, noise, scaling, and large angle rotation operations.

In today's world of fast communication, to insure, the security and integrity of information is a big challenge. Data hiding also known as Steganography is one of the fields that deal in methods related information security and hide secret information and message other information. This paper elucidates the effect of data hiding in different number of least significant bits in the primary colors of RGB color image. These individual color channels are analyzed at different hiding capacity level and its has been observed that high quality Stego images with PNSR 30dB and above can been obtained by hiding secret information in 5 least significant bits of red, green and blue channels, specially the green and blue channels give a very high visual quality. The individual channels can hide 20% data, i.e. one fifth of the overall size of cover image, with undetectable changes in cover image.

In order to evaluate the complexity of a color image more effectively and find the connection between image complexity and image information, this paper presents a method to compute the complexity of image based on color information.Under the complexity ,the theoretical analysis first divides the complexity from the subjective level, divides into three levels: low complexity, medium complexity and high complexity, and then carries on the image feature extraction, finally establishes the function between the complexity value and the color characteristic model. The experimental results show that this kind of evaluation method can objectively reconstruct the complexity of the image from the image feature research. The experimental results obtained by the method of this paper are in good agreement with the results of human visual perception complexity,Color image complexity has a certain reference value.

Studies have been actively undertaken on image retrieval from a large-capacity image database, particularly on image retrievals such as natural image retrieval, character image retrieval or trademark retrieval from a streaming image database. In retrieving a character image, either color or shape information is used for the key feature information. However, changes in the shape of a character image often makes it difficult to retrieve solely based on the shape information, requiring a new retrieval method where both color and shape information are taken into consideration. We present a highly effective method for retrieving or matching similar character images even when shape information differs substantially. In our approach, combined features of color and shape information are used for image retrieval; an image is first split into sectors; color information is extracted from the sector images followed by quantization using Parzen window to extract features; an image is then retrieved by means of bipartite matching using the features. Our results show the retrieval rate using the combined information increases substantially for matching natural or character images compared with the results obtained by the combination of two features independently.

Tristimulus display of the abundant information contained in a hyper spectral image is a challenging task. Previous visualization approaches focused on preserving as much information as possible in the reduced spectral space, but ended up with displaying hyper spectral images as false color images, which contradicts with human experience and expectation. This paper proposes a new framework to tackle this problem. It is based on the fusion of a hyper spectral image and a high-resolution color image via manifold alignment technique. Manifold learning is an important tool for dimension reduction. Manifold alignment projects a pair of two data sets into a common embedding space so that the pairs of corresponding points in these two data sets are pair wise aligned in this new space. Hyper spectral image and high-resolution color image have strong complementary properties due to the high spectral resolution in the former and the high spatial resolution in the latter. The embedding space produced by manifold alignment bridges a gap between the high dimensional spectral space of hyper spectral image and RGB space of color image, making it possible to transfer the natural color and spatial information of a high-resolution color image to a hyper spectral image to generate a visualized image with natural color distribution and finer details.

A persistent problem in the three-dimensional (3-D) reconstruction technique is to eliminate blank areas in the 3-D map, which commonly emerges after removing undesired objects, such as dynamic targets or occluded areas. This task is challenging for it is difficult to acquire the coherence between color and depth information, which are both lost for each pixel in the target region. Moreover, creating novel artifacts also needs to be avoided during the completion process. To address these problems, in this paper, we propose a collaborative method to complete the lost area in the color image and its corresponding depth map. In the proposed method, an examplar-based image inpainting technique combined with planarity knowledge is adopted to iteratively repair the texture and structure information. The color completion process provides candidates for plane-fitting on segmented superpixels. The depth filling step extracts image areas requiring reinpainting, which is defined by their gradient differences. The final 3-D map can then be reconstructed from both completed color and depth images. Experiments on real-world scenes demonstrate the success of our method in the completion of 3-D maps.

In this work, we treat the problem of combined classification of a high spatial resolution color image and a lower spatial resolution hyperspectral image of the same scene. The problem is particularly challenging, since we aim for classification maps at the spatial resolution of the color image. Contextual information is obtained from the color image by introducing Color Attribute Profiles (CAPs). Instead of treating the `R', `G', and `B' bands separately, the color image is transformed into CIE-Lab space. In this color space, attribute profiles are extracted from the `L' band, which corresponds to the Luminance, while the `a' and `b' bands, which contain the color information, are kept intact, and the resulting images are transformed back into RGB space. The spectral information is obtained from the hyperspectral image. A Composite Decision Fusion (CDF) strategy is proposed, combining a state-of-the-art kernel-based decision fusion technique with the popular composite kernel classification approach. Experiments are conducted, using simulated data and a real multisource dataset containing airborne hyperspectral data and orthophotographic data from a suburban area in Belgium. These experiments show that our CAPs perform well with respect to other approaches for extracting attribute profiles from high resolution color images, and that the proposed CDF strategy produces meaningful results with respect to concatenation and the highlighted state-of-the-art approaches for combining multisource data.