Dynamic fusion of images from the visible and infrared channels of sightseeing system by complex matrix formalism

Author(s):  
D. Khaustov ◽  
Ya. Khaustov ◽  
Ye. Ryzhov ◽  
O. Burashnikov ◽  
E. Lychkovskyy ◽  
...  

The employment of new mathematical and computer approaches for the fusion of target images from the visible and infrared channels of the sightseeing system (SSS) is one of the ways to increase the efficiency of the SSS of armored vehicles. Modern approaches to improving the efficiency of image fusion are aimed to increase the visibility of the target via improving the quality indices of fused images. This paper proposes a fundamentally new approach to image fusion, namely dynamic image fusion, at which the target is observed in the mode of a video clip composed of a sequence of stationary fused images obtained at different parameters of fusion, in contrast to traditional stationary image fusion, at which the decision is made from one fused image. Unlike stationary image fusion, aimed to increase the visibility of the target, the dynamic image fusion allows one to enhance the conspicuity of the target. The principle of dynamic image fusion proposed in this paper is based on matrix formalism, in which the fused image is constructed in the form of a complex vector function, which by its mathematical form is analogous to the Jones vector of elliptically polarized light wave, which in turn opens the possibility of matrix transformation of the complex vector of the fused image and consequently its parameterization by analogy with the Jones matrix formalism for the light wave. The article presents mathematical principles of matrix formalism, which is the basis for dynamic image fusion, gives examples of stationary and dynamic image fusion by the method of complex vector function and compares with the corresponding images, fused by algorithms of weight addition in the field of real and complex scalars. It is shown that by selecting weight coefficients, the general form of a complex amplitude vector image can be reduced to the algorithms of weight and averaged addition in the field of real scalars, weight amplitude and RMS-image in the field of complex scalar numbers, and geometric-mean image in the field of complex vectors, which, thereby, are partial cases of the general form of the complex amplitude image in the field of complex vectors. The animated images obtained by the method of complex vector function illustrate the increase of conspicuity of the object of observation due to the dynamic change of the fusion parameters.

Image Fusion ◽  
2020 ◽  
pp. 171-296
Author(s):  
Gang Xiao ◽  
Durga Prasad Bavirisetti ◽  
Gang Liu ◽  
Xingchen Zhang

Author(s):  
Vladimir Petrovic ◽  
Tim Cootes ◽  
Rade Pavlovic

Author(s):  
Zhe Wang

Art design is based on art and images, in which dynamic images require pre-processing and fusion. This paper explores how to apply the dynamic image fusion technology in the development of the innovation capabilities of college students majoring in art design. The research results show that image sequences with different exposures can be directly formed into low dynamic range images through multi-exposure fusion, or the high dynamic range can be restored first, and then converted to a low dynamic range image through tone mapping. The main principles for innovation development in art design education include conformity, student-centeredness, and overall optimization. The dynamic image fusion enhancement algorithm is mainly divided into five functional modules, which are used to calculate multi-scale gradient, structure tensor, and target gradient for the fused image, solve the steepest descent method and output normalized images, respectively


2007 ◽  
Vol 8 (1) ◽  
pp. 56-62 ◽  
Author(s):  
Zhong-hua Wang ◽  
Zheng Qin ◽  
Yu Liu

Sensors ◽  
2021 ◽  
Vol 21 (3) ◽  
pp. 863
Author(s):  
Vidas Raudonis ◽  
Agne Paulauskaite-Taraseviciene ◽  
Kristina Sutiene

Background: Cell detection and counting is of essential importance in evaluating the quality of early-stage embryo. Full automation of this process remains a challenging task due to different cell size, shape, the presence of incomplete cell boundaries, partially or fully overlapping cells. Moreover, the algorithm to be developed should process a large number of image data of different quality in a reasonable amount of time. Methods: Multi-focus image fusion approach based on deep learning U-Net architecture is proposed in the paper, which allows reducing the amount of data up to 7 times without losing spectral information required for embryo enhancement in the microscopic image. Results: The experiment includes the visual and quantitative analysis by estimating the image similarity metrics and processing times, which is compared to the results achieved by two wellknown techniques—Inverse Laplacian Pyramid Transform and Enhanced Correlation Coefficient Maximization. Conclusion: Comparatively, the image fusion time is substantially improved for different image resolutions, whilst ensuring the high quality of the fused image.


Sign in / Sign up

Export Citation Format

Share Document