A digital tobacco leaf and a manufacturing method thereof

Abstract

The purpose of the present invention is to provide a digital tobacco leaf and a production method thereof, which relates to the technical field of cigarette production. The digital tobacco leaf is composed of image information of a standard tobacco leaf, and the image information at least includes original spectral image data of the standard tobacco leaf, data of a non-green tobacco area, and data of a green tobacco area. The digital tobacco leaf proposed by the present invention can save spectral images of tobacco leaves of various grades to quickly evaluate the grade of the tobacco leaves.

Description

Technical Field

[0001] The present invention relates to the technical field of cigarette production, and in particular to a digital tobacco leaf and a production method thereof. Background Art

[0002] In tobacco leaf production, in order to distinguish tobacco leaves of different qualities and make each grade of tobacco leaves have relatively consistent quality for subsequent cigarette production, it is necessary to grade tobacco leaves. Among them, the appearance quality factor is an important grading factor for measuring tobacco leaf grades.

[0003] In the prior art, the evaluation of tobacco leaf appearance quality is mainly done manually, that is, the grade and quality of tobacco leaves are determined by touch and visual inspection. In tobacco leaf grading, the color of tobacco leaves is an important evaluation indicator of their quality grade. Among them, the basic colors of tobacco leaves include lemon yellow, orange yellow, and red brown; the non-basic colors of tobacco leaves include green-yellow, slightly green, and variegated colors. The accurate identification of the above colors requires a high level of quality from the evaluation experts, and is easily interfered by various external factors in practical applications.

[0004] In view of this, the inventor, based on many years of experience in production design in this field and related fields, has designed a digital tobacco leaf and a production method thereof through repeated experiments in order to solve the problems existing in the prior art. Summary of the invention

[0005] The object of the present invention is to provide a digitized tobacco leaf and a production method thereof, which can store spectral images of tobacco leaves of various grades to quickly evaluate the grade of the tobacco leaves.

[0006] To achieve the above-mentioned object, the present invention proposes a digitized tobacco leaf, wherein the digitized tobacco leaf is composed of image information of a standard tobacco leaf, and the image information at least includes original spectral image data of the standard tobacco leaf, data of a non-green tobacco area, and data of a green tobacco area.

[0007] The present invention also proposes a method for producing digitized tobacco leaves, wherein original spectral image data of standard tobacco leaves are collected, blue smoke area data and non-blue smoke area data of the standard tobacco leaves are obtained through the original spectral image data, and the original spectral image data, the blue smoke area data and the non-blue smoke area data are stored as the digitized tobacco leaves.

[0008] Compared with the prior art, the present invention has the following characteristics and advantages:

[0009] The digital tobacco leaves and the production method thereof proposed in the present invention can save the spectral image information of the standard tobacco leaves through the digital tobacco leaves, and can truly display the appearance characteristics of the standard tobacco leaves. As a digital form of the standard tobacco leaves, the digital tobacco leaves can not only more accurately and quickly evaluate the grade of the tobacco leaves to be tested, but also can be used as a reference at various stages of the cigarette production process, and provide a basis for the subsequent automation and digitalization improvements in cigarette production. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The accompanying drawings described herein are only for explanation purposes and are not intended to limit the scope of the present invention in any way. In addition, the shapes and proportional dimensions of the various components in the figures are only schematic, used to help understand the present invention, and are not specifically limited to the shapes and proportional dimensions of the various components of the present invention. Those skilled in the art can select various possible shapes and proportional dimensions to implement the present invention according to the teachings of the present invention.

[0011] Figure 1 is an image of a standard tobacco leaf in the present invention;

[0012] Figure 2 The present invention provides a flow chart of a method for making digital tobacco leaves.

[0013] Figure 3 is an original spectral image of a standard tobacco leaf in one embodiment of the present invention;

[0014] Figure 4 The image is a result of identification of a standard tobacco leaf with a slight green band in one embodiment of the present invention;

[0015] Figure 5 is an original spectral image of a standard tobacco leaf in another embodiment of the present invention;

[0016] Figure 6 The image is a result of identification of a standard tobacco leaf with a slight green band in another embodiment of the present invention; DETAILED DESCRIPTION

[0017] The details of the present invention can be more clearly understood with reference to the accompanying drawings and the description of the specific embodiments of the present invention. However, the specific embodiments of the present invention described herein are only used for the purpose of explaining the present invention and cannot be understood as limiting the present invention in any way. Under the guidance of the present invention, technicians can conceive of any possible variations based on the present invention, which should all be considered to belong to the scope of the present invention.

[0018] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it may be directly on the other element or there may be an intermediate element. When an element is referred to as being "connected to" another element, it may be directly connected to the other element or there may be an intermediate element.

[0019] The present invention provides a digitized tobacco leaf, which is composed of image information of a standard tobacco leaf, wherein the image information at least includes original spectral image data of the standard tobacco leaf, data of a non-green tobacco area, and data of a green tobacco area.

[0020] The present invention also provides a method for making a digital tobacco leaf, such as Figures 2 to 6 As shown, original spectral image data of standard tobacco leaves are collected, blue smoke area data and non-blue smoke area data of the standard tobacco leaves are obtained through the original spectral image data, and the original spectral image data, the blue smoke area data and the non-blue smoke area data are stored as the digitized tobacco leaves.

[0021] The digital tobacco leaves and the production method thereof proposed in the present invention can save the spectral image information of the standard tobacco leaves through the digital tobacco leaves, and can truly display the appearance characteristics of the standard tobacco leaves. As a digital form of the standard tobacco leaves, the digital tobacco leaves can not only more accurately and quickly evaluate the grade of the tobacco leaves to be tested, but also can be used as a reference at various stages of the cigarette production process, and provide a basis for the subsequent automation and digitalization improvements in cigarette production.

[0022] In an optional embodiment of the present invention, the image information also includes the green tobacco distribution ratio of the standard tobacco leaf.

[0023] In an optional embodiment of the present invention, Figure 1 As shown, standard tobacco leaves are selected according to the tobacco grade standards of various countries to meet the corresponding cigarette production standards.

[0024] In an optional example of this embodiment, the upper limit standard of standard tobacco leaves is fresh tobacco leaves.

[0025] In an optional example of this embodiment, the lower limit standard of the standard tobacco leaves is normal tobacco leaves that do not contain green tobacco.

[0026] In an optional embodiment of the present invention, the step of obtaining the blue tobacco area data includes preprocessing the original spectral image to obtain preprocessed spectral information, obtaining the spectral curve of each pixel on the leaf surface of the tobacco leaf to be evaluated through the preprocessed spectral information, calculating the curvature radius of the spectral curve of each pixel, extracting the blue tobacco pixels according to each curvature radius and displaying them as a blue tobacco area.

[0027] In an optional embodiment of the present invention, the original spectral image of the standard tobacco leaf is collected by a hyperspectral imager.

[0028] Hyperspectral imagers can use existing technologies. By measuring the radiation energy emitted or reflected by the sample, they can obtain images of dozens to hundreds of bands within a narrow band range, can achieve object classification and detection, and can organically combine traditional two-dimensional imaging technology and spectral technology. They have spectral analysis and image processing capabilities, and have high spectral resolution, which can obtain rich information. By obtaining detailed spectral feature information and external characteristic information of the sample through the hyperspectral imager, a more accurate and comprehensive evaluation of the measured substance can be made.

[0029] In an optional example of the present invention, the model of the hyperspectral imaging spectrometer is GaiaField-V10E-AZ4; further, the recognition unit also includes a CCD camera, four 50W halogen lamps, and an electric displacement control platform.

[0030] When acquiring the original image, the standard tobacco leaves are first placed on the electric displacement control platform, and then the movement of the electric displacement control platform and the acquisition of the hyperspectral image are controlled by the computing unit. Finally, the acquired hyperspectral image data is transmitted to the computing unit (computer) for storage.

[0031] In an optional example, when a hyperspectral imager is used to collect original spectral images of tobacco leaves, standard tobacco leaves are fully unfolded manually and placed on the stage of the hyperspectral imager, and the spectral information of the front and back sides of the standard tobacco leaves are collected respectively, and saved according to digital numbers.

[0032] In an optional example of this embodiment, the preprocessing process includes using a standard black and white plate correction formula to perform data correction on the original atlas information.

[0033] Furthermore, the standard black and white plate correction formula is:

[0034]

[0035] In the formula, the standard white calibration plate obtains the all-white calibration image W; the all-black calibration image B is obtained by scanning the built-in blackboard, the original image I is corrected, and the preprocessed atlas data is R.

[0036] Specifically, before collecting the sample to be tested, it is necessary to collect the standard white calibration plate and the built-in blackboard to calibrate the collected hyperspectral image in order to eliminate the influence of instrument noise. W is a fully white calibration image obtained by using the standard white calibration plate. The whiteboard is moved under the camera to collect and save its hyperspectral image; B is a fully black calibration image obtained by scanning the built-in blackboard. The camera lens is covered with a camera cover to collect and save the fully black hyperspectral image; I is the sample to be tested placed on the electric displacement control platform for collection; R is the spectral image obtained after the instrument noise is eliminated and smoothed after calibration with the black and white plates.

[0037] In an optional embodiment of the present invention, obtaining the spectral value of each pixel on the leaf surface of the tobacco leaf to be identified through the preprocessed spectral information includes:

[0038] The preprocessed spectral information is converted into single-band spectrum data; the single-band spectrum data is divided into tobacco leaf information and background information; the tobacco leaf information is divided into leaf surface information, leaf stem information and wrinkle information, and the leaf surface information is marked as a sensing area; the spectrum of the area of ​​interest is corrected, the leaf surface information within a predetermined band range is intercepted, and the curvature radius of the spectral curve of each pixel in the leaf surface information is calculated one by one.

[0039] In an optional example of this implementation, the spectrum information at a wavelength of 675.4 nm in the pre-processed spectral information is selected and converted into binary single-band spectrum data.

[0040] In an optional example of this embodiment, a global threshold method is used to segment tobacco leaves and background information, and then a local threshold method is used to segment leaf surface, leaf stem and wrinkle information, and the leaf surface portion is marked as a region of interest.

[0041] Among them, the threshold segmentation method of the image is intuitive and easy to implement, and it plays an important role in the application of image segmentation. Specifically, the image f(m,n) consists of two types of regions with different gray levels: dark objects and bright objects. The bright and dark parts of this image can be clearly distinguished in the histogram, so a threshold can be selected to separate the bright and dark peaks. Threshold segmentation can be regarded as a function operation:

[0042] T = T[m,n,p(m,n),f(m,n)];

[0043] Where m and n represent the horizontal and vertical coordinates of the pixel, p(m, n) represents the local characteristics of the pixel, and f(m, n) represents the pixel gray value; the image after thresholding is defined as:

[0044]

[0045] Among them, the pixels marked with grayscale 1 correspond to the object of interest, and the pixels marked with 0 correspond to the background, and there are:

[0046] 1. If T depends only on,f(m,n), the threshold is global;

[0047] 2. If T depends on,f(m,n) and p(m,n), the threshold is local;

[0048] 3. If T depends on the coordinates (m, n), the threshold is adaptive.

[0049] In the above process, the global threshold can be understood as counting the histogram of the entire image, and then segmenting all the pixels in the image by selecting a certain threshold; while the local threshold is to divide the image into set small blocks, count the histogram in the small blocks, select a certain threshold, and segment the pixels in the small blocks; the local threshold segmentation is more detailed.

[0050] In an optional embodiment of the present invention, the radius of curvature is calculated using the formula:

[0051]

[0052] In the formula, x is the spectral band value, and y is the reflectance of the corresponding spectral band value.

[0053] In an optional example, the spectrum of the region of interest is first corrected using a multivariate scattering correction method, and then the 600nm-700nm band is intercepted to calculate the radius of curvature of the spectrum curve of each pixel one by one.

[0054] The multivariate scattering correction method corrects the baseline shift and offset of spectral data by averaging the spectrum, which can effectively eliminate the spectral differences caused by different scattering levels, thereby enhancing the correlation between the spectrum and the data.

[0055] The specific steps of the multivariate scattering correction method are as follows: (1) Calculate the average value of all spectral data; (2) Perform a univariate linear regression between the spectrum of each sample and the average spectrum, and solve the least squares problem to obtain the baseline shift and offset of each sample; (3) Correct the spectrum of each sample by subtracting the obtained baseline shift and dividing by the offset to obtain the corrected spectrum.

[0056] In an optional implementation manner of the present invention, the process of extracting corresponding pixels according to each of the curvature radii and displaying them as blue smoke regions includes:

[0057] The pixels are sorted from low to high according to the radius of curvature, and then the radius of curvature is normalized, and the quantile is used as an indicator to set a heat map to display the distribution of the green smoke area on the leaf surface.

[0058] Normalization is to convert data into decimals in the range (0,1); to convert dimensional expressions into dimensionless expressions to ensure data comparability.

[0059] In an optional example of this embodiment, the quantile method is used to divide the radius of curvature of each pixel after sorting into four equal parts, and the quantiles for identifying blue tobacco are 10%, 20%, and 30% of the radius of curvature according to the numerical value. The quantile method can also use other percentages. The selection of different quantiles will result in different distribution effects of blue tobacco, which can be understood as display areas with different degrees of blue. If the set quantile is large, the blue area will be displayed more completely, and there may be partial misidentification; if the set quantile is small, the blue area will be displayed as an area with a larger degree of blue; and the quantile selected in this embodiment is divided with reference to the GB 2635-92 standard for flue-cured tobacco, and what is finally displayed is the pixel area that is smaller than the set quantile value.

[0060] In an optional embodiment of the present invention, the production method further includes calculating a green smoke distribution ratio, where the green smoke distribution ratio is the ratio of the number of pixels in the green smoke area to the number of pixels in the tobacco leaf.

[0061] It should be noted that the spectral curve of leaves with different degrees of green will have absorption peaks with different degrees of curvature, and this feature is the characteristic peak of the green part; at the same time, the radius of curvature is the main method used to measure the degree of curvature of the curve. The radius of curvature can better fit the characteristic peak. The radius of curvature is a measure without the need for prior knowledge learning, avoiding the trouble of manual data annotation, and can automatically perform rapid identification.

[0062] The detailed explanation of the above-mentioned embodiments is only intended to explain the present invention so as to facilitate a better understanding of the present invention. However, these descriptions cannot be interpreted as limitations on the present invention for any reason. In particular, the various features described in different embodiments may also be arbitrarily combined with each other to form other embodiments. Unless there is a clear description to the contrary, these features should be understood to be applicable to any embodiment and are not limited to the described embodiments.

Claims

1. A method for making digital tobacco leaves, characterized in that: The original spectral image data of the standard tobacco leaf is collected, and the blue smoke area data and non-blue smoke area data of the standard tobacco leaf are obtained through the original spectral image data, and the original spectral image data, the blue smoke area data and the non-blue smoke area data are stored as the digitized tobacco leaf; the step of obtaining the blue smoke area data includes preprocessing the original spectral image to obtain preprocessed spectral information, obtaining the spectral curve of each pixel of the leaf surface of the tobacco leaf to be evaluated through the preprocessed spectral information, calculating the curvature radius of the spectral curve of each pixel, extracting the blue smoke pixel according to each curvature radius and displaying it as the blue smoke area; Obtaining the spectral curve of each pixel on the leaf surface of the tobacco leaf to be evaluated through the preprocessed spectral information includes: Converting the preprocessed spectral information into single-band spectrum data; segmenting the single-band spectrum data into tobacco leaf information and background information; segmenting the tobacco leaf information into leaf surface information, leaf stem information and wrinkle information, and marking the leaf surface information as a region of interest; correcting the spectrum of the region of interest, intercepting the leaf surface information within a predetermined band, and calculating the curvature radius of the spectral curve of each pixel in the leaf surface information one by one; The radius of curvature is calculated using the formula: ， （2） In the formula, ; , x is the spectral band value, y is the reflectance of the corresponding spectral band value; The process of extracting corresponding pixels according to the curvature radius and displaying them as smoke areas includes: The pixels are sorted from low to high according to the radius of curvature, and then the radius of curvature is normalized, and the quantile is used as an indicator to set a heat map to display the distribution of the green smoke area on the leaf surface.

2. The method for making digital tobacco leaves according to claim 1, characterized in that: The preprocessing process includes using a standard black and white plate correction formula to perform data correction on the original atlas information.

3. The method for making digital tobacco leaves according to claim 2, characterized in that: The standard black and white plate correction formula: ， （1） In the formula, the standard white calibration plate obtains the all-white calibration image W; the all-black calibration image B is obtained by scanning the built-in blackboard, the original image I is corrected, and the preprocessed atlas data is R.

4. The method for making a digitized tobacco leaf according to claim 1, characterized in that: The production method also includes calculating the green smoke distribution ratio, where the green smoke distribution ratio is the ratio of the number of pixels in the green smoke area to the number of pixels in the tobacco leaves.

Images