TWI391872B - Multi - band image imaging method - Google Patents

Description

Multi-band image imaging method

The present invention relates to an imaging method, and more particularly to an imaging method for multi-band imaging.

Since the early observation of gastric lesions in patients with metal tubes, endoscopy has been booming in the diagnosis and treatment of hollow organs. In order to improve the correctness of diagnosis and improve the efficiency of treatment, the technology of improving endoscopes, whether it is the structure of endoscopes or the function of endoscopes, has been progressing year by year. The treatment technology of endoscopes Early development of Endoscopic Mucosal Resection (EMR), and the development of endoscopic Submucosal Dissection (ESD), has been developed to reduce the operation of hollow organs. Treatment, and avoid the inconvenience caused by surgical treatment, because today's endoscopic submucosal separation can be used to remove the organ tissue of malignant lesions, so that patients with malignant lesions can be exempted from general surgical treatment, and malignant lesions are removed. .

For the judgment of the lesion, after the lesion is found under the endoscope, it is still necessary to judge the external shape of the lesion according to the endoscopic image obtained by the endoscope, for example, the contour, the size and the like. Please refer to the first figure, which is a flow chart of a conventional intraocular image imaging method. As shown in the figure, the imaging method of the digital intra-view image is used to image the image data generated by the image sensing unit of the endoscope. First, as shown in step S10, the image sensing unit is read. The image data is continued as shown in step S12, and the image processing operation is performed on the read image data to calculate the image feature recorded by the image data, and then step S14 is performed to generate an image according to the image feature calculated in step S12. Corresponding to the image data generated by the image sensing unit of the endoscope, finally, as shown in step S16, the image generated in step S14 is displayed. Thus, the doctor obtains the endoscopic image for diagnosis or treatment by using the image data generated by the endoscope, but the imaging method of the conventional endoscopic image cannot clearly display the external features of the organ tissue, for example, the lesion on the organ tissue The edge position.

Once the diagnosis is performed by using an endoscope, it is necessary to slice the organ tissue to detect the threat of the lesion by using the slice of the organ tissue. However, the imaging method of the conventional endoscopic image cannot clearly show the organ tissue. The feature makes the slicing process easy to fail, and the slice of the healthy organ tissue often causes the doctor to fail the slice processing, and the diagnosis of the lesion produces a wrong judgment.

Therefore, how to solve the above problems and propose a multi-band image imaging method can not only improve the shortcomings of the traditional endoscopic image, but also clearly show the lesions, and improve the diagnostic efficiency, and can solve the above problems.

The main object of the present invention is to provide a multi-band image imaging method, which uses a multi-band image data as a plurality of bands of image data and reconstructs different band image data to clearly display the organ tissue shape by different band image data. feature.

The present invention is a multi-band image imaging method, which is applied to an image sensing unit for sensing a multi-band image, and generates an image data according to the multi-band image sensed by the image sensing unit, wherein the image data includes a plurality of bands. image. The imaging method of the invention first reads out the image data, and then separates and processes the image data to be separated, and separates the image data into multiple bands, and then reconstructs the image data of the band to separate the separated image data. The image data of the band is restored to the complete band image data, and then the image is generated according to the reconstructed band image data to display the image corresponding to the image data of different bands on the external display device. Since the images produced by different band image data have different effects on displaying the shape characteristics of the organ tissues, the present invention can generate corresponding images according to different needs for diagnosis and treatment.

For a better understanding and understanding of the structural features and the efficacies of the present invention, please refer to the preferred embodiments and the detailed descriptions as follows: please refer to the second figure, which is A flow chart of an embodiment of the invention. As shown, the present invention The invention relates to a multi-band image imaging method, which is applied to process image data generated by an image sensing unit according to a multi-band image, wherein the image data comprises a plurality of bands of image data. The imaging method of the present invention is initially as shown in step S20, and the image sensing unit senses the multi-band image to generate the image data, and subsequently reads the image data as shown in step S22. This embodiment is Reading the image data to an image processing platform, and then separating according to the band image data included in the image data, as shown in step S24, for reading the band image data in the subsequent step, The embodiment separates a narrow-band image and a wide-band image, wherein the narrow-band image and the wide-band image have different bandwidths, and the narrow-band image is a single-band image, such as a single band of RGB or a CMY color model. The blue light image is composed of a plurality of wavelength bands, for example, a white light image, which can be composed of RGB or three primary color lights of a CMY color model. In addition, the image data can also separate a plurality of narrow-band image data. That is, the multiple single-band image data is separated.

Receiving the above, the narrow-band image data is read as shown in step S26, and the wide-band image data is read as shown in step S28 at the same time; thereafter, the read in step S26 is reconstructed as shown in step S30. The narrow-band image data is reconstructed by a bilinear interpolation method to reconstruct the narrow-band image, and the wide-band image data read in step S28 is reconstructed as shown in step S32. Bilinear interpolation to reconstruct the wideband image. In addition, the reconstruction operation of the present embodiment uses a bilinear interpolation method. In addition, the reconstruction operation of the present invention may also employ a Laplaciam Bilinear interpolation method or other image processing method. Finally, as shown in step S34 and step S36, the narrow-band image and the wide-band image reconstructed in step S30 and step S32 are simultaneously displayed, wherein the image characteristics displayed by the narrow-band image are different from the wide-band image, so that the width can be widened. Differences between band images and narrow-band images, as well as medical diagnosis and observation.

The image data captured by the conventional image sensing unit is the third A picture. The image sensing unit of the present invention senses the image data as shown in the third B picture. In step S24, the third B image data is separated into complex band image data, as shown in the third C to the third D. The third to third D pictures of the embodiment adopt the gradation values of the three primary colors of RGB, and in addition, the color gradation values of the CMY color model or the gradation values of other color models may be used to represent the color, the third A The figure shows a multi-band image sensed by a conventional image sensing unit, wherein each square represents each recorded in the image data. One pixel, and each pixel includes the color gradation of the three primary colors of RGB, and the third B picture is the image data generated by the image sensing unit of the present invention, each square only stores a single-band gradation value, and after separation In the image data, the image data corresponding to the narrow-band image includes a single-band data as shown in the third C-picture; and the image data corresponding to the wide-band image, as shown in the third D-picture, includes a plurality of different-band data.

In step S30, the narrow-band image reconstruction method uses the bilinear interpolation method to interpolate the narrow-band image read in step S26, as shown in the fourth to fourth B-pictures. The fourth A picture to the fourth B picture are gradation values of the three primary colors of RGB, in addition, the color value of the CMY color model or the color gradation value of other color models may also be used to represent the color, wherein the fourth A picture is not The reconstructed image data, NB12, NB14, NB32, NB34, NB52 and NB54 are the gradation values of the narrow band color, and the fourth B picture is the image data during reconstruction, and the reconstructed complete image data, as shown in the sixth figure . The equation for the reconstruction operation of the narrow-band image is taken as an example of the nine-pixel pixel of the fourth B-picture, wherein X13, X22, X23, X24, and X33 are the gradation values obtained by the pixel reconstruction operation, and the operation equation is as follows. :X13=(NB12+NB14)/2,X22=(NB12+NB32)/2,X24=(NB14+NB34)/2,X23=(NB12+NB14+NB32+NB34)/4,X33=(NB32+ NB34)/2.

In step S32, the wideband image is also subjected to bilinear interpolation to interpolate the wideband image read in step S30, as shown in the fifth A to fifth B, wherein the fifth A is For unreconstructed image data, R11, R13, R15, R31, R33, R35, R51, R53 and R55 are red gradation values, B21, B23, B25, B41, B43 and B45 are blue gradation values, G22, G24 G42 and G44 are green gradation values, and the fifth B is the image data during reconstruction, and the reconstructed complete image data is shown in the seventh figure. The reconstruction method of the wide-band image is calculated by the nine-square grid method, in which X12, X14, X32, X34, X52 and X54 are the gradation values of the blanks, and the pixels are interpolated with the gradation values of other colors, for example: The pixel where X32 is located is interpolated R32 and G32. With the color gradation value of B32, R32 is (R31+R33)/2, G32 is (G22+G24)/2, B32 is (B21+B23+B41+B43)/4, and the pixel system where R33 is located is interpolated with G33 and The color gradation value of B33, G33 is (G22+G24+G42+G44)/4, B33 is (B23+B43)/2, the pixel of B23 is the gradation value of interpolation R23 and G23, and R23 is (R13+R33) ) /2, G23 is (G22 + G24) / 2, the above calculation method is to use R33 as the center point of the nine squares and interpolate with the gradation values of the surrounding pixels.

In steps S32 and S36, the narrow-band image and the wide-band image are as shown in the sixth and seventh figures, and the images in the sixth and seventh images are the same scene, but the wide-band image shown in the seventh figure The contour of the lesion cannot be clearly displayed, and the distribution of the microvessels is also unclear, but the narrow-band image of the sixth image is not only obvious for the contour of the lesion, but also for the narrow-wavelength images for microvessels and other finer image features. It can be clearly stated that it is beneficial for the user to use the reconstructed wide-band image to reconstruct the narrow-band image for research reference or diagnostic treatment.

In summary, the present invention is a multi-band image imaging method for processing multi-band image data generated by a multi-band image sensing unit, which first separates different band image data to make different band images. After the reconstruction operation, the images of different bands can be obtained for the user to compare and judge the difference according to the image of different bands, so that different image features can be presented for the same scene, which is beneficial for observation or research reference.

Therefore, the present invention is a novelty, progressive and available for industrial use. It should be in accordance with the patent application requirements stipulated in the Patent Law of China, and the invention patent application is filed according to law, and the prayer bureau will grant the patent as soon as possible. For prayer.

However, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention, and the shapes, structures, features, and spirits described in the claims are equivalently changed. Modifications are intended to be included in the scope of the patent application of the present invention.

The first figure is a flow chart of a conventional intraocular image imaging method; the second figure is a flow chart of an embodiment of the present invention; and the third A is a schematic view of a conventional image data band; The third B is a schematic diagram of the un-separated band of the image data of the present invention; the fourth A is a schematic diagram of the un-reconstructed narrow-band image data of the present invention; and the fourth B is a schematic diagram of the narrow-band image data reconstruction of the present invention. 5A is a schematic diagram of the wide-band image data of the present invention being unreconstructed; FIG. 5B is a schematic diagram of the wide-band image data reconstruction of the present invention; and FIG. 6 is a schematic diagram of the narrow-band image of the present invention; The seven figures are schematic diagrams of the wide band image of the present invention.

Claims (16)

A multi-band image imaging method for processing a multi-band image data generated by a multi-band image sensing unit, the multi-band image data comprising a plurality of band images, the imaging method comprising: reading the multi-band image data The plurality of pixels of the multi-band image data respectively have color gradation values of different bands; the multi-band images are separated according to the multi-band image data; and the integrity interpolation operation is performed on the complex pixel data of the same band to reconstruct the same band The plurality of pixel data is a complete band image data; and the plurality of band image images are generated according to the reconstructed band image complete data. The imaging method of claim 1, wherein the step of separating the multi-band image data according to the image data further comprises a step of reading the multi-band image data. The imaging method of claim 2, wherein the step of reading the multi-band image data further comprises a step of reading out image data of different bands. The imaging method of claim 3, wherein in the step of reconstructing the multi-band image data, the reconstructing the different band image data. The imaging method of claim 3, wherein the different band image data comprises at least one narrow band image data. The imaging method of claim 5, wherein the narrowband image data comprises a plurality of pixels of a monochrome image. The imaging method of claim 6, wherein the monochrome image is a color corresponding to RGB or a CMY color model. The imaging method of claim 3, wherein the different band image data comprises a narrow band image data and a wide band image data. The imaging method of claim 8, wherein the narrow-band image data is a plurality of pixels of a monochrome image. The imaging method of claim 9, wherein the monochrome image is a color corresponding to RGB or a CMY color model. The imaging method of claim 8, wherein the broadband image data is a white light image data. The imaging method of claim 11, wherein the white light image data comprises three primary colors of RGB or a CMY color model, and the plurality of pixels of the white light image data respectively are color gradation values corresponding to different primary colors. The imaging method of claim 1, wherein after the step of generating the plurality of images based on the reconstructed plurality of bands of image data, the method further comprises a step of displaying the images. The imaging method according to claim 1, wherein in the step of reconstructing the image data of the band, the method of reconstructing the band images by using a bilinear interpolation method or adopting a pull conversion interpolation method ( Laplaciam Bilinear interpolation) reconstructs the band images for integrity interpolation operations on complex pixel data in the same band. The imaging method of claim 1, wherein the image sensing unit is a CMOS transistor sensor. The imaging method of claim 1, wherein the image sensing unit is a CCD sensor.