Conventional RGB image acquisition employs an image capturing system that utilizes color filter array (CFA) technology; however, it is limited in its ability to represent visible colors. Multispectral imaging based on filter array architecture is required for general image capturing because of its exquisite color representation system. However, it has several issues associated with spatial and spectral resolution.In this paper, we propose a new demosaicking method that improves reconstructed image quality by considering interchannel correlation. Our proposed method strengthens the crosscorrelation of demosaicked channels by repeating interpolations. Experimental results show that our proposed method generates better quality reconstructed images than conventional methods.Index Terms-multispectral filter array, multispectral demosaicking, color difference, inter-channel correlation, spectral correlation I. INTRODUCTION Multispectral imaging utilizes more spectra than RGB imaging to precisely represent colors for various computer vision applications, medical imaging, and satellite imaging.Multispectral images are captured by several systems [1]: (i) multi-camera-multi-shot systems, (ii) single-camera-multishot systems, and (iii) single-camera-one-shot systems. Multicamera-multi-shot systems consist of an optical splitter and a number of RGB cameras with different color filters. Although these systems can also capture images at video rate by using RGB video capturing systems in each camera, the equipment used in these systems may be quite expensive. Single-cameramulti-shot systems can comprehensively capture all pixel values in all bands by sequentially replacing filters in front of the camera. However, these cameras cannot capture images at video rate because of the speed of sequentially replacing color filters. Single-camera-one-shot systems can observe several color bands by using color filter array (CFA) in the RGB domain, and also multispectral filter array (MSFA) [2] in the multispectral domain. These systems generate profit at low cost, and capture images at video rate.Conventional RGB cameras commonly adopt singlecamera-one-shot techniques. These cameras consist of a single image sensor with CFA and can observe several color bands by observing one band at each pixel of a single image sensor, with the observed band being determined by the pattern of CFA. Thus, full color images are estimated from observed data captured by CFA. This process of estimation is called demosaicking.
Hyperspectral imaging (HSI) provides more detailed information than red-green-blue (RGB) imaging, and therefore has potential applications in computer-aided pathological diagnosis. This study aimed to develop a pattern recognition method based on HSI, called hyperspectral analysis of pathological slides based on stain spectrum (HAPSS), to detect cancers in hematoxylin and eosin-stained pathological slides of pancreatic tumors. The samples, comprising hyperspectral cubes of 420-750 nm, were harvested for HSI and tissue microarray (TMA) analysis. As a result of conducting HAPSS experiments with a support vector machine (SVM) classifier, we obtained maximal accuracy of 94%, a 14% improvement over the widely used RGB images. Thus, HAPSS is a suitable method to automatically detect tumors in pathological slides of the pancreas.
A capturing system with multispectral filter array (MSFA) technology is proposed for shortening the capture time and reducing costs. Therein, a mosaicked image captured using an MSFA is demosaicked to reconstruct multispectral images (MSIs). Joint optimization of the spectral sensitivity of the MSFAs and demosaicking is considered, and pathology-specific multispectral imaging is proposed. This optimizes the MSFA and the demosaicking matrix by minimizing the reconstruction error between the training data of a hematoxylin and eosin-stained pathological tissue and a demosaicked MSI using a cost function. Initially, the spectral sensitivity of the filter array is set randomly and the mosaicked image is obtained from the training data. Subsequently, a reconstructed image is obtained using Wiener estimation. To minimize the reconstruction error, the spectral sensitivity of the filter array and the Wiener estimation matrix are optimized iteratively through an interior-point approach. The effectiveness of the proposed MSFA and demosaicking is demonstrated by comparing the recovered spectrum and RGB image with those obtained using a conventional method.
Mosaicked color filter arrays and demosaicking methods for multispectral images have been proposed in recent years. Several studies have evaluated the multispectral filter array (MSFA) pattern, but a method to optimize the pattern has not been proposed. We focus on the spatial arrangement of MSFAs to improve the demosaicked image quality. To evaluate the filter arrangement, we propose a new metric that uses both the spatial and spectral correlation of the filters. The arrangement is optimized by using simulated annealing and the proposed metric. An advantage of the proposed metric and its optimization is that they do not need to use image data. Experimental results show that the new metric is proportional to the peak-to-signal noise ratio (PSNR) and that a higher PSNR can be obtained by minimizing the metric using simulated annealing.
A new filter array and a demosaicking method for snapshot multispectral polarization imaging are proposed in this paper. The proposed filter array is a thin-film wavy multilayer structure regarded as a photonic crystal that can be fabricated using the autocloning method. The multispectral polarization filter array is developed by altering the wave structure of the photonic crystal at each pixel. In addition, we propose a demosaicking method for multispectral polarization images by considering snapshot imaging as a linear model. In the experiments, we evaluated the recovered spectrum error in some color charts and showed various demosaicked images such as multispectral polarization images, specific-band degree of linear polarization images, polarized RGB images, and non-polarized RGB images.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.