Unlock instant, AI-driven research and patent intelligence for your innovation.
Tobacco leaf part discrimination method based on spectral characteristic parameter fusion and probability classification
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A spectral feature and probability classification technology, applied in the field of near-infrared non-destructive testing, can solve the problem of low identification rate of tobacco leaf parts, and achieve the effect of improving classification accuracy, increasing discrimination, and rapid part analysis
Pending Publication Date: 2022-02-11
CHINA TOBACCO ZHEJIANG IND +1
View PDF3 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0067] Step 1. Sample making
[0068] A total of 8511 spectral data of raw tobacco samples of various grades were collected in Yunnan, Hunan, and Guizhou from 2018 to 2020, and N=8511. Spectral data with a waveband of 892-1699nm was selected, and the number of wavebands L=257 was obtained by interpolation. Mark the sample part information, the upper part is marked as U, the middle part is marked as C, and the lower part is marked as D. image 3 Spectral curves of some samples are given.
[0069] Step 2. Fusion of spectral characteristic parameters
[0070] Spectral characteristic parameters take the first-order differential spectrum.
[0071] For each sample V i (i=1,2,…,N), calculate its first order differential spectrum
[0072]
[0073] where v ij for V i The jth element of , L is the number of spectral bands;
[0074] obtained by normalization Then get a 1×512 sample vector by splicing
[0075]
[0076] Figure 4 The spectral curves of some sample data ...
Embodiment 2
[0087] Steps 1 to 3 of Embodiment 2 are exactly the same as Embodiment 1.
[0088] Step 4, identification of tobacco leaf parts:
[0089] Tobacco leaf part discrimination adopts incomplete determination method, and the specific process is as follows:
[0090] (1) Obtain the naturalization vector according to the classification probability
[0091]
[0092]
[0093] the p 1i = u i ,p 2i =c i ,p 3i = d i ,q ij is the normalized vector q i The jth column element of .
[0094] (2) Judging the tobacco leaf position according to the value of the classification probability normalization vector
[0095]
[0096] where Y i Indicates the location of the i-th sample.
[0097] Figure 5 The comparison of the recognition rate of tobacco leaf parts based on SVM probability classification based on the original data and the fusion data of spectral feature parameters is given, where Figure 5 (a) Tobacco leaf part discrimination adopts incomplete deterministic method, 5(...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
PUM
Login to View More
Abstract
The invention discloses a tobacco leaf part distinguishing method and device based on near infrared spectrum characteristic parameter fusion and probability classification. The method comprises steps of collecting a sample spectrum on a near infrared spectrometer, and marking the category of the tobacco leaf part; spectral characteristic parameters of the collected sample data being extracted, and the characteristic parameters and original data being spliced and fused; randomly selecting a certain proportion of fusion data as a training data set of a probability classifier, and selecting an optimal parameter of the classifier through a verification curve and training; the spectral data of any tobacco leaf sample being inputted into the trained classifier after being subjected to feature parameter fusion, and the classification probability that the tobacco leaf sample belongs to the upper tobacco, the middle tobacco and the lower tobacco being outputted; and obtaining a normalized vector according to the classification probability, and judging whether the tobacco leaf part is a deterministic part or a transitional part according to the value of the normalized vector. According to the method, an efficient and stable technical means is provided for lossless judgment of tobacco leaf parts by utilizing a near infraredspectrum analysis technology.
Description
technical field [0001] The invention belongs to the technical field of near-infrared non-destructive testing, and in particular relates to a tobacco leaf part discrimination method and device based on near-infrared spectrum feature parameter fusion and probability classification. Background technique [0002] Because tobacco leaves from different places of origin and different parts have different smoke characteristics, the place of origin and part of the tobacco leaves should be marked when storing in the warehouse. The reason makes it impossible to judge the origin and location of the tobacco leaves, which greatly affects the subsequent processing and manufacturing of tobacco leaves. [0003] The traditional origin and identification of tobacco leaf parts mainly rely on manual sensory evaluation methods, that is, experts in the tobacco industry use their senses of vision, smell, touch and taste to measure. This method will consume a lot of manpower, material and financial ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
Application Information
Patent Timeline
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More