RU2692575C1 - Method of integrating digital halftone images - Google Patents

Abstract

FIELD: calculating; counting.SUBSTANCE: invention relates to the computer equipment. Method of integrating two digital halftone images includes determining the most informative image, forming the resulting image and aligning its brightness range, wherein, when integrating images, calculating differences of brightness values of each pixel of the most informative image with surrounding pixels, obtained differences are averaged, complexed halftone image is formed by summation of brightness values of pixels of the second image with corresponding values of average difference of pixels of the most informative image with given gain factor, resultant image is represented in the form of RGB-colour image with three components: two initial and complexed images.EFFECT: technical result is higher local contrast of resulting image.1 cl, 11 dwg

Description

The invention relates to the field of digital image processing, in particular to the integration of digital images.

The prior art knows how to integrate digital halftone images obtained from two channels of different spectral range (Patent RU 2451338 published 05/20/2012, IPC G06T 5/00). The presented method consists in receiving the original television and thermal images, determining the main image with the highest level of informative details, calculating the average brightness of the image of the second channel, calculating the average value of the absolute differences between the average brightness of the image of the second channel and the brightness values of all the pixels of the image of the second channel, for each pixel of the integrated image, calculate the sum of the brightness value of the pixel of the image of the main channel and the absolute the difference between the brightness values of the pixel image of the second channel and the average brightness of the image of the second channel, subtract the average absolute values of the brightness differences between the images of the second channel image and the average brightness of the second channel image from the sum, normalize the brightness range of the image.

One of the disadvantages of this method is the reduction of the local contrast of objects, especially in the case of the presence of highlights on one of the original images.

The closest to the proposed method is the integration of two digital half-tone images (Patent RU 2540778 published 02.10.2015, IPC G06T 5/50, G06K 9/46). This method includes obtaining original images, determining the most informative image, combining images based on a combination of the most informative image and the distinctive details of the second image, leveling the brightness range of the resulting image. When complexing images, estimates of linear regression coefficients of brightness values of the most informative image by brightness values of the second image are determined, an intermediate image is formed as a result of a linear transformation of the most informative image with found coefficients, a combined image is found by adding the difference between the second and intermediate images to the most informative image.

The disadvantages of the prototype method is to reduce the local contrast of objects in the resulting image as in the presence of highlights on one of the original images, and in the case of similar object brightness and background spectral images, as well as the difficulty of identifying spectrally dependent objects on the resulting complex image.

The technical result of the proposed method of integrating digital halftone images is to increase the local contrast of the resulting image containing the elements of the original images of the same scene obtained in different spectral ranges, as well as simplifying the detection of spectrally dependent objects.

The technical result is achieved by the fact that according to the proposed method including determining the most informative image, forming the resulting image and aligning its brightness range, when combining images, calculate the difference in brightness values of each pixel of the most informative image with surrounding pixels, average the resulting differences, form an integrated halftone image by summing the brightness values of the pixels of the second image with the corresponding conductive average difference values most informative image pixels with a predetermined gain, the resulting image is presented in the form of RGB-color image with three components: two initial images and complexed image.

The essence of the proposed method lies in the fact that when integrating images, calculate the difference between the brightness values of each pixel of the most informative image (for example, a television (TV) image) with the surrounding pixels:

;where i = 1, ..., m; j = 1, ..., n; m, n - the number of rows and columns of the TV image

;

p = 0, ..., P is the parameter that determines the coordinates of the neighborhood pixel by i,

q = 0, ..., Q is the parameter that determines the coordinates of the neighborhood pixel by j.

d = 1, ..., ((2Р + 1) (2Q + 1)) - 1 - index of the difference between the brightness values of i, j-th pixel with the pixels of the neighborhood.

Average the obtained difference by the number of pixels in the neighborhood:

Form a complex halftone image by summing the brightness values of the pixels of the second image (for example, thermal imaging (TFT) image) with the corresponding values of the average difference in brightness of the pixels of the most informative image with a given gain:

where k is the gain of the average difference in brightness of the pixels of the most informative image.

с тремя компонентами (например

:According to (see, for example, J. Agoston. Theory of Color and its Application in Art and Design: Trans. From English - M .: Mir, 1985. 184), for visualizing the distinctive features of multi-spectral images, the resulting image is represented as a synthesized RGB color image

with three components (for example

:

Align the luminance range of the resulting color image using any known method (see, for example, Gonzalez R., Woods R. Digital image processing. M .: Tekhnosfera, 2005. 1072 sec.).

Display the resulting color image for viewing.

The proposed method can be implemented, for example, using a device whose block diagram is shown in FIG. 1. The block diagram of the device contains: 1 - a block of preliminary preparation of images that performs geometric transformation of images, improving brightness and contrast, filtering noise, 2 - a block for selecting the most informative image; 3 - a unit for calculating the difference in brightness values of each pixel of the most informative image with its surrounding pixels; 4 - a unit for calculating the average value of the difference in pixel brightness of the most informative image; 5 is a block to enhance the average values of the difference in pixel brightness of the most informative image; 6 - block the formation of the integrated halftone image; 7 is a block forming the resulting RGB-color complexed image; 8 - block alignment of the luminance range of the color image; 9 - output unit of the resulting color image.

The device in FIG. 1 works as follows. Two multi-spectral images are input to the device, which are fed to the input of the preliminary image preparation block 1, then the images are fed to the block of selection of the most informative image 2, after which the most informative image is fed in parallel to the block for calculating differences in brightness values of each pixel of the most informative image pixels 3 and the block forming the resulting RGB-color complex image 7, and the second image in parallel serves on the block forming a complexed halftone image 6 and a block forming the resulting RGB-color complexed image 7. After processing in block 3, the resulting differences in the brightness values of each pixel of the most informative image with surrounding pixels go to the unit calculating the average value of the brightness differences of pixels in the most informative image 4, after what the obtained average values of the difference in brightness of the pixels of the most informative image is fed to the unit of formation of the set xenated halftone image 6, the resulting integrated halftone image is fed to the forming unit of the resulting RGB color integrated image 7. The resulting RGB color integrated image is fed to the brightness leveling unit of the color image 8, after which the resulting image is fed to the output unit of the resulting color image 9.

Consider examples. FIG. Figures 2, 7 show the original TV image (in Fig. 2, the illumination of objects in the scene, in Fig. 7 poorly visible objects with low contrast). FIG. 3, 8 shows the original TPR image with contrasting objects. FIG. 4, 9 shows the result of the integration of the original multi-spectral half-tone images into a single half-tone image according to the prototype method. FIG. 5, 10 presents an integrated halftone image obtained on the basis of the proposed method. FIG. 6, 11 in order to demonstrate the achieved technical result, the resulting color integrated image obtained by the image complexing device (Fig. 1) is presented on the basis of the proposed method, in which objects of both the first and second source images are clearly visible, as well as spectrally highlighted in color dependent objects and scene elements.

Comparing the results of complexing multi-spectral halftone images known (Fig. 4, 9) and the proposed (Fig. 6, 11) method, you can see that the inventive method allows to get the best result, because the resulting integrated image is more informative with high local contrast and allows selection of spectrally dependent objects according to their color difference, both in conditions of glare interference, and in conditions of low contrast of one of the original images.

Thus, the proposed method allows to increase the local contrast of informative elements and objects, as in the case of the presence of highlights on one of the source images, and in the case of close object brightness and background of different spectral images, as well as to simplify the identification of spectrally dependent objects, due to the fact that when complexing images, calculate the difference of the brightness values of each pixel of the most informative image with the surrounding pixels, average the obtained differences, form a complex A halftone image by summing the brightness values of the pixels of the second image with the corresponding values of the average pixel difference of the most informative image with a given gain, the resulting image is represented as an RGB-color image with three components: two source images and an integrated image. Thus, the proposed method eliminates the disadvantages of the prototype.

The proposed technical solution is new, because of the publicly available information is unknown method of integrating digital halftone images, which consists in determining the most informative image, forming the resulting image and aligning its brightness range, when combining images calculating the difference in brightness values of each pixel of the most informative image with its surrounding pixels, averaging the obtained differences, the formation of an integrated halftone image by summing the brightness values of the pixels of the second image with the corresponding values of the average pixel difference of the most informative image with a given gain, representing the resulting image as an RGB-color image with three components: two source images and an integrated image.

The proposed technical solution is industrially applicable, since its implementation can be used any known from the prior art programmable and non-programmable processors digital signal processing and images (see, for example, URL: http://module.ru/catalog/).

Claims (1)

The method of integrating two digital halftone images, including determining the most informative image, generating the resulting image and aligning its luminance range, characterized in that when combining images, calculate the difference between the brightness values of each pixel of the most informative image and the surrounding pixels, averaged halftone image by summing the brightness values of the pixels of the second image with tvetstvuyuschimi average difference values most informative image pixels with a predetermined gain, the resulting image is presented in the form of RGB-color image with three components: two initial images and complexed image.

Images