Visual identification method and system for stem cut tobacco

Abstract

The invention discloses a stem sliver cut tobacco visual identification method and system, and the method comprises the steps: collecting an image of a cut stem to be identified, and obtaining a cut stem image, the cut stem comprising a stem sliver and / or cut tobacco; based on the cut stem image, extracting cut stem features; and obtaining a visual identification result of the cut stem image by using the cut stem features and a pre-constructed visual identification model. According to the stem and tobacco shred visual identification method and system, automatic identification of the stem and the tobacco shred can be achieved, the stem and the tobacco shred can be accurately distinguished and removed, the precision of identifying the tobacco shred and the stem can reach about 98%, the accuracy and intelligent degree of identification and metering can be improved, and the stem rejection rate can be calculated more rapidly and accurately on the basis.

Description

Technical Field

[0001] The present invention relates to the technical field of tobacco processing, and more specifically, to a method and system for visually identifying stemmed tobacco. Background Art

[0002] The control of stem shreds sorting in the cigarette making machine is an important part of quality and consumption management. The tobacco stems that cannot be completely removed from the tobacco sheets form a small amount of stem sticks after the shredding and drying stages. The hard and sharp stem sticks seriously affect the sensory and smoking quality of the finished cigarettes. The cigarette making machine uses a wind selection device to separate the tobacco shreds and stem sticks before the cigarettes are formed, which is the last step in the removal of stem sticks. The separated stem sticks are generally mixed with a small amount of tobacco shreds. If the stem stick removal rate is too low, the stem stick content of the cigarette will be too high, which may cause the weight deviation of the cigarette to increase, increase the probability of defects such as cigarette puncture, damage the combustion performance and sensory quality of the cigarette, and affect the stability of product quality; if the stem stick removal rate is too high, the tobacco shreds in the removed stem sticks will be too high, the tobacco consumption of the cigarette making machine will increase, and the effective utilization rate of tobacco shreds will be reduced.

[0003] At present, after the air selection process, cigarette factories mainly use manual methods to identify stems and tobacco, and then calculate the stem rejection rate. The manual identification method is inefficient, the consistency of identification results from different people is poor, and the identification accuracy needs to be improved.

[0004] Therefore, there is an urgent need for a method and system for visually identifying stemmed tobacco. Summary of the invention

[0005] The object of the present invention is to provide a method and system for visually identifying stemmed tobacco shreds to solve the problems in the above-mentioned prior art, to realize automatic identification of stem shreds and tobacco shreds, to accurately distinguish between stemmed shreds and tobacco shreds, and to improve the accuracy and intelligence of identification and measurement.

[0006] The present invention provides a method for visually identifying stemmed tobacco, which comprises the following steps:

[0007] Collecting an image of the cut stems to be identified to obtain a cut stem image, wherein the cut stems include stem sticks and / or cut tobacco;

[0008] Extracting stem features based on the stem image;

[0009] The visual recognition result of the cut stem image is obtained by using the cut stem features and a pre-built visual recognition model.

[0010] In the above-mentioned method for visually identifying stemmed tobacco, preferably, the step of collecting an image of the stems to be identified to obtain the stem image specifically includes:

[0011] The image of the shredded stem to be identified is collected by a high-definition camera to obtain the shredded stem image.

[0012] In the above-mentioned method for visually identifying stemmed tobacco, preferably, the stem features include at least one of shape features, size features and color features.

[0013] In the above-mentioned method for visually identifying stemmed tobacco, preferably, the extracting stem features based on the stem image specifically includes:

[0014] intercepting the outline of each material in the shredded stem image;

[0015] The contour of each material in the shredded stem image is input into a pre-trained shredded stem feature extraction model to obtain shredded stem features.

[0016] In the above-mentioned method for visually identifying stem tobacco, preferably, the stem feature extraction model includes an image classification network.

[0017] In the above-mentioned method for visually identifying stemmed tobacco, preferably, the step of obtaining the visual recognition result of the stem image by using the stem features and a pre-built visual recognition model specifically includes:

[0018] The shredded stem features are input into a pre-built visual recognition model to obtain a visual recognition result of the shredded stem image.

[0019] In the above-mentioned method for visually identifying stem tobacco, preferably, the visual recognition model includes a visual recognition model constructed based on a convolutional neural network.

[0020] The above-mentioned method for visually identifying stemmed tobacco, wherein preferably, the method for visually identifying stemmed tobacco further comprises:

[0021] The stem shreds image is preprocessed.

[0022] In the above-mentioned method for visually identifying stemmed tobacco, preferably, the preprocessing of the stem image specifically includes:

[0023] Performing filtering on the stem image to remove noise in the image;

[0024] The filtered stem image is binarized to enhance the contrast in the image.

[0025] The present invention also provides a tobacco stem visual recognition system using the above method, comprising:

[0026] A cut stem image acquisition module is used to acquire an image of the cut stems to be identified to obtain a cut stem image, wherein the cut stems include stem sticks and / or cut tobacco;

[0027] A stem cut feature extraction module, used for extracting stem cut features based on the stem cut image;

[0028] The visual recognition module is used to obtain a visual recognition result of the cut stem image by using the cut stem features and a pre-built visual recognition model.

[0029] The present invention provides a method and system for visually identifying stem-cut tobacco. Based on a stem-cut image, stem-cut features are extracted, and then the visual recognition result of the stem-cut image is obtained by using the stem-cut features and a pre-built visual recognition model, so as to realize automatic recognition of stem-cut and tobacco-cut, accurately distinguish stem-cut and tobacco-cut, and achieve an accuracy of about 98% in identifying tobacco-cut and stem-cut. The accuracy and intelligence of recognition and measurement can be improved, so as to facilitate more rapid and accurate calculation of the stem-cut rejection rate on this basis. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described below with reference to the accompanying drawings, in which:

[0031] Figure 1 A flow chart of an embodiment of a method for visually identifying stemmed tobacco provided by the present invention;

[0032] Figure 2 This is a structural block diagram of an embodiment of a tobacco stem visual recognition system provided by the present invention. DETAILED DESCRIPTION

[0033] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. The description of the exemplary embodiments is merely illustrative and is in no way intended to limit the present disclosure and its application or use. The present disclosure can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to make the present disclosure thorough and complete and to fully express the scope of the present disclosure to those skilled in the art. It should be noted that unless otherwise specifically stated, the relative arrangement of the components and steps, the composition of the materials, the numerical expressions and the numerical values ​​set forth in these embodiments should be interpreted as being merely exemplary and not as limiting.

[0034] The words "first", "second" and similar words used in this disclosure do not indicate any order, quantity or importance, but are only used to distinguish different parts. The words "include" or "comprises" and similar words mean that the elements before the word include the elements listed after the word, and do not exclude the possibility of including other elements. "Up", "down" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0035] In the present disclosure, when a specific component is described as being located between a first component and a second component, there may or may not be an intermediate component between the specific component and the first component or the second component. When a specific component is described as being connected to other components, the specific component may be directly connected to the other components without an intermediate component, or may not be directly connected to the other components but have an intermediate component.

[0036] All terms (including technical terms or scientific terms) used in the present disclosure have the same meanings as those understood by ordinary technicians in the field to which the present disclosure belongs, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries, such as general dictionaries, should be interpreted as having meanings consistent with their meanings in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined herein.

[0037] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0038] like Figure 1 As shown, the actual implementation of the method for visually identifying stemmed tobacco provided in this embodiment specifically includes the following steps:

[0039] Step S1, collecting an image of shredded stems to be identified to obtain an image of shredded stems, wherein the shredded stems include stem sticks and / or shredded tobacco.

[0040] Specifically, the image of the shredded tobacco stems to be identified is collected by a high-definition camera to obtain the shredded tobacco stem image. In one embodiment of the present invention, a high-speed line scan CCD industrial camera is used to collect the image of the shredded tobacco stems to be identified. In a specific implementation, the shredded tobacco stems can be laid loosely and flatly to completely separate the stems and shredded tobacco, and then the shredded tobacco stem image is collected by a high-definition camera. The high-definition camera can ensure that a high-quality shredded tobacco stem image is obtained.

[0041] Furthermore, in some embodiments of the present invention, between step S1 and step S2, the method for visually identifying stemmed tobacco may further include:

[0042] Step S1-2, preprocessing the cut stem image.

[0043] Image quality can be improved by image preprocessing. In one embodiment of the method for visually identifying stemmed tobacco of the present invention, step S1-2 may specifically include:

[0044] Step S1-2-1, filtering the stem image to remove noise in the image.

[0045] Step S1-2-2: binarize the filtered stem image to enhance the contrast of the image.

[0046] Among them, the stem image can be binarized by using the threshold segmentation method to facilitate the detection of the image edge.

[0047] Step S2: extracting the features of the shredded stems based on the shredded stems image.

[0048] The shredded stems characteristics include at least one of shape characteristics, size characteristics and color characteristics.

[0049] In one embodiment of the method for visually identifying stemmed tobacco of the present invention, step S2 may specifically include:

[0050] Step S21, capturing the outline of each strand of material in the shredded stem image.

[0051] Step S22: inputting the outline of each piece of material in the shredded stem image into a pre-trained stem feature extraction model to obtain shredded stem features.

[0052] The stem feature extraction model includes an image classification network. By extracting the stem features, it is possible to distinguish them from other impurities.

[0053] Step S3: using the shredded stem features and a pre-built visual recognition model to obtain a visual recognition result of the shredded stem image.

[0054] Specifically, the shredded tobacco stem features are input into a pre-constructed visual recognition model to obtain a visual recognition result of the shredded tobacco stem image. By recognizing the shredded tobacco stem image through the visual recognition model, it is possible to accurately distinguish the de-stemmed tobacco from the shredded tobacco, and to improve the accuracy and intelligence of recognition and measurement. In one embodiment of the present invention, the visual recognition model includes a visual recognition model constructed based on a convolutional neural network. It should be noted that the present invention does not specifically limit the type of the visual recognition model, and in other embodiments of the present invention, the visual recognition model may also be other types of deep learning models.

[0055] In order to verify the recognition accuracy of the visual recognition model for stems and shredded tobacco, a certain amount of pure shredded tobacco and pure stems were tested to observe the detection effect. The test results are shown in the table.

[0056] Table 1 Detection effect of visual recognition model on stems and tobacco

[0057] As can be seen from Table 1, the accuracy of the present invention in identifying tobacco shreds and stems can reach about 98%.

[0058] The method for visually identifying stemmed tobacco provided by the embodiment of the present invention extracts stem features based on a stem image, and then uses the stem features and a pre-built visual recognition model to obtain a visual recognition result of the stem image, thereby realizing automatic identification of stems and tobacco, and can accurately distinguish between stemmed tobacco and tobacco. The accuracy of identifying tobacco and stemmed tobacco can reach about 98%, which can improve the accuracy and intelligence of identification and measurement, and facilitates faster and more accurate calculation of the stem rejection rate on this basis.

[0059] Correspondingly, if Figure 2 As shown, the present invention also provides a tobacco stem visual recognition system, comprising:

[0060] The cut stem image acquisition module 101 is used to acquire an image of the cut stems to be identified, and obtain a cut stem image, wherein the cut stems include stem sticks and / or cut tobacco;

[0061] A stem cut feature extraction module 102, configured to extract stem cut features based on the stem cut image;

[0062] The visual recognition module 103 is used to obtain a visual recognition result of the cut stem image by using the cut stem features and a pre-built visual recognition model.

[0063] In the visual recognition system for stem and shredded tobacco provided by the embodiment of the present invention, the stem feature extraction module extracts stem features based on the stem image, and the visual recognition module uses the stem features and a pre-built visual recognition model to obtain a visual recognition result of the stem image, thereby realizing automatic recognition of stems and shredded tobacco, and can accurately distinguish between stem-removed stems and shredded tobacco, with an accuracy of about 98% in recognizing shredded tobacco and stems, thereby improving the accuracy and intelligence of recognition and measurement, and facilitating faster and more accurate calculation of the stem rejection rate on this basis.

[0064] So far, various embodiments of the present disclosure have been described in detail. In order to avoid obscuring the concept of the present disclosure, some details known in the art are not described. Based on the above description, those skilled in the art can fully understand how to implement the technical solution disclosed here.

[0065] Although some specific embodiments of the present disclosure have been described in detail by way of examples, it should be understood by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present disclosure. It should be understood by those skilled in the art that the above embodiments may be modified or some technical features may be replaced by equivalents without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims

1. A method for visually identifying stemmed tobacco, characterized in that: include: Collecting an image of the cut stems to be identified to obtain a cut stem image, wherein the cut stems include stem sticks and / or cut tobacco; Extracting stem features based on the stem image; The visual recognition result of the cut stem image is obtained by using the cut stem features and a pre-built visual recognition model.

2. The method for visually identifying stemmed tobacco according to claim 1, characterized in that: The collecting of the image of the shredded stems to be identified to obtain the shredded stems image specifically includes: The image of the shredded stem to be identified is collected by a high-definition camera to obtain a shredded stem image.

3. The method for visually identifying stemmed tobacco according to claim 1, characterized in that: The cut stem characteristics include at least one of shape characteristics, size characteristics and color characteristics.

4. The method for visually identifying stemmed tobacco according to claim 1, characterized in that: The extracting the stem features based on the stem image specifically includes: intercepting the outline of each material in the shredded stem image; The contour of each material in the shredded stem image is input into a pre-trained shredded stem feature extraction model to obtain shredded stem features.

5. The method for visually identifying stemmed tobacco according to claim 1, characterized in that: The stem signature feature extraction model includes an image classification network.

6. The method for visually identifying stemmed tobacco according to claim 1, characterized in that: The step of obtaining a visual recognition result of the cut stem image by using the cut stem feature and a pre-built visual recognition model specifically includes: The shredded stem features are input into a pre-built visual recognition model to obtain a visual recognition result of the shredded stem image.

7. The method for visually identifying stemmed tobacco according to claim 1, characterized in that: The visual recognition model includes a visual recognition model constructed based on a convolutional neural network.

8. The method for visually identifying stemmed tobacco according to claim 1, characterized in that: The method for visually identifying stemmed tobacco also includes: The stem shreds image is preprocessed.

9. The method for visually identifying stemmed tobacco according to claim 8, characterized in that: The preprocessing of the stem image specifically includes: Performing filtering on the stem image to remove noise in the image; The filtered stem image is binarized to enhance the contrast in the image.

10. A tobacco stem visual recognition system using the method according to any one of claims 1 to 9, characterized in that: include: A cut stem image acquisition module is used to acquire an image of the cut stems to be identified to obtain a cut stem image, wherein the cut stems include stem sticks and / or cut tobacco; A stem feature extraction module, used for extracting stem features based on the stem image; The visual recognition module is used to obtain a visual recognition result of the cut stem image by using the cut stem features and a pre-built visual recognition model.

Images