CN114743082A - Cigarette finished product cigarette box characteristic identification method based on machine vision - Google Patents
Cigarette finished product cigarette box characteristic identification method based on machine vision Download PDFInfo
- Publication number
- CN114743082A CN114743082A CN202210194180.6A CN202210194180A CN114743082A CN 114743082 A CN114743082 A CN 114743082A CN 202210194180 A CN202210194180 A CN 202210194180A CN 114743082 A CN114743082 A CN 114743082A
- Authority
- CN
- China
- Prior art keywords
- cigarette
- stack
- goods
- steps
- layer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02W—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO WASTEWATER TREATMENT OR WASTE MANAGEMENT
- Y02W90/00—Enabling technologies or technologies with a potential or indirect contribution to greenhouse gas [GHG] emissions mitigation
Landscapes
- Wrapping Of Specific Fragile Articles (AREA)
Abstract
The invention discloses a cigarette finished product cigarette box characteristic identification method based on machine vision, which belongs to the technical field of intelligent logistics, and is characterized in that an intelligent camera is used for identifying whether a goods position of goods in a warehouse is provided with the goods, whether a film is wrapped, a product gauge is adopted, whether a cigarette box is recycled, the number of the cigarette box and the content requirements of products in production/finished products are met, and the content requirements are compared with inventory information to complete warehouse checking work. The method has good accuracy and stability, and solves the potential safety hazard problem of manual checking; the checking efficiency is improved, the manpower resource is saved, and unmanned inventory checking is realized; and the real-time image of the cigarette in the checking process is stored to prove the authenticity and effectiveness of the physical checking.
Description
Technical Field
The invention relates to the technical field of intelligent logistics, in particular to a cigarette finished product cigarette box characteristic identification method based on machine vision.
Background
Finished cigarette products produced by tobacco group cigarette factories are stacked on trays in a cigarette box mode according to specified standards and stored in an elevated warehouse, and the storage management inventory checking mode adopts artificial vision to identify, so that the problems of potential safety hazards of overhead operation, low efficiency, accuracy, timeliness and the like exist. With the development of an image recognition technology and an automation technology, an automatic warehouse inventory solving idea based on machine vision is provided, an intelligent camera is used for photographing the characteristics of each cigarette box in a warehouse, the images are analyzed through algorithm software, the identification requirements of whether the goods position of the goods in the warehouse is a real object, whether a film is wrapped, whether the goods is a standard, whether the cigarette box is recovered, the number of stack-type cigarette boxes, the products in process/finished products and the like are identified, and the identification requirements are compared with inventory information to finish warehouse inventory work. The stacker can continuously work without human interference in the checking process, and the visual recognition system records pictures of each tray warehouse, so that the tray tracing and manual correction operations are realized.
And performing characteristic analysis on different identification requirements of the finished cigarette box, selecting one or more characteristics for each identification requirement as an identification algorithm, performing characteristic comparison with a large number of photographed pictures, and adjusting the characteristic selection of different identification requirements according to the identification result of the identification algorithm on the large number of pictures.
Elevated library environment:
(1) the goods positions of the cigarette finished product elevated warehouse are defined according to the position of row-column-layer, the space size of each goods position is uniform, after the cigarette boxes forming a stack on the tray are put in storage, the residual space is small, and the installation position of the camera is limited;
(2) the background of the empty goods space is complex, the influence of open ambient light is obvious, and the interference is large;
(3) the cigarette brands are of various types, and the series sub-brands are very small in characteristic difference on appearance and easy to confuse;
(4) some solid tray cigarette stacks are used for preventing the cigarette pieces from falling off, and part of cigarette box group trays are wrapped with films (containing various cigarette piece stack types), so that the characteristics of the cigarette boxes can be covered;
(5) the cigarette carton after the group plate is placed in different postures, some front faces (Chinese characters) are in front, some back faces (letters) are in front, some back faces are in upside down (the front and the back of the pattern are used as differences), and the same brand feature model is built more.
The accuracy of the recognition requirements and statistics including whether the goods location is provided with or not, whether the goods location is wrapped with a film or not, the goods gauge, the stack type, whether the bar code is provided or not, the identification of the recycling box, the number of cigarette pieces and the like needs to be considered:
(1) how to accurately identify targets at high speed (in real time) for a stack of multiple brands of cartons;
(2) how to effectively construct and organize a reliable recognition algorithm and to smoothly implement;
(3) real-time is an important issue that is difficult to solve. The image processing speed and the algorithm running speed are one of the main bottlenecks affecting the real-time performance of a visual system;
(4) stability is a first consideration for vision systems, and the following problems are faced with either location-based, image-based, or hybrid visual servoing methods: when the image features are very small in difference, how to combine the global features with the local features is considered for ensuring the stability of the system, namely, the feature region is enhanced and the global convergence is ensured; in order to avoid the failure of single algorithm identification, consideration needs to be given to how to adopt other supplementary algorithms to ensure stability and accuracy.
Disclosure of Invention
The invention aims to solve the problems, realize machine vision inventory and provide a cigarette finished product cigarette box characteristic identification method based on machine vision. The following technical scheme is adopted:
in order to realize the purpose, the invention is realized by adopting the following technical scheme: the method comprises a method for identifying whether a goods location has a real object or not; a method for identifying whether the tobacco stack is wrapped with a film; identifying the product gauge; a method of identifying whether to recover the smoke box; a method for identifying the number of the stack-type smoke boxes; method for identification of articles in manufacture/finished products.
Preferably, the method requires the use of stock smoke box features including: the goods position comprises a cigarette stack, empty goods position inventory information, tray group inventory information, a reflection light spot after the cigarette stack is wrapped with a film, characters, patterns, colors and bar codes on the cigarette box, a white mark pasted on the recovered cigarette box, a mark on a fixed position, a cigarette box stack type main view, a top view and a first cigarette box engineering bar code.
Preferably, the method for identifying whether the goods space has a real object comprises the following steps: the method comprises the steps of extracting colors of various cigarette boxes when a cigarette stack exists in a goods position, establishing a model, searching colors corresponding to the model in a region of interest (ROI), and if the colors are found to indicate that the cigarette stack exists in the goods position, otherwise, the colors are empty goods positions, the effect of identifying whether the goods position is a real object or not is greatly influenced by images of the empty goods positions and a tray group, and empty goods position inventory information and tray group inventory information are needed to improve identification accuracy.
Preferably, the method for identifying whether the cigarette stack is wrapped with the film comprises the following steps: after the cigarette stack is wrapped with the film, the collected image has bright light spots under the angle of illumination light, according to the characteristic, a gray level histogram analysis algorithm is adopted, the color image is converted into a gray level image point by point in an interested area, then the gray level histogram structural data of the area is calculated and analyzed, and whether the wrapping film exists or not is determined by adopting the contrast in the structural data.
Preferably, the method for identifying the specification comprises the following steps:
pattern training in the step (1), which comprises the following steps: selecting training patterns, setting a training area and an original point, setting training parameters, training the patterns, evaluating trained characteristics, and training a pattern model according to the principle: selecting a representative pattern having consistent characteristics; unnecessary features and image noise are reduced; training only important features;
step (2), an algorithm during operation: setting parameters during operation, defining a search area, acquiring patterns during operation, operating an intelligent camera algorithm and acquiring a result.
Preferably, the method for identifying whether to recover the smoke box comprises the following steps: the mark of the recycling bin is marked by pasting a white mark on the used recycling bin and writing numbers at a fixed position in the production process;
step (1), establishing a white-mark pattern model by adopting a pattern matching algorithm, searching a matched target in an image, and if the matched target is 1, identifying the target and recovering the smoke box; if the number is 0, the cigarette is not marked, and the cigarette is a new cigarette box;
and (2) reading in a fixed ROI by adopting an OCR character recognition algorithm in a mode of writing numbers in a fixed position, and reading out the number of the numbers which is the recycling times of the smoke box.
Preferably, the method for identifying the number of the stack-type cigarette boxes comprises the following steps:
step (1) collecting model images of each stack shape and each position, and carrying out image processing;
step (2) establishing a main view model (full stack, each layer) and a overlook model (each layer and each position);
step (3) real-time image processing of the cargo space;
step (4) judging whether the primary view is full of stacks, if so, obtaining corresponding stack types and total number according to a primary view full stack model, and if not, entering the next step;
step (5) judging whether the main view has four layers, if so, counting the number of the fourth layer plus the number of the first three layers according to the main view, and if not, entering the next step;
step (6) judging whether the main view has three layers or not, if so, counting the number of the third layer plus the number of the first two layers according to the main view, and if not, entering the next step;
step (7) judging whether the main view has two layers, if so, counting the number of the second layer plus the number of the previous layer according to the top view, and if not, entering the next step;
and (8) judging whether the main view is a layer or not, counting the number of the first layer according to the top view if the main view is the layer, and judging the goods position or the tray group to be empty if the main view is the layer.
Preferably, the identification method of the products under production/finished products comprises the following steps: the finished product has the first project code according to the fact that the finished product has no first project code. The identification algorithm of products in process is that firstly, a pattern model of the products in process is established, the position of a target is found in an image, then a bar code of the position is read, and a plurality of results are subjected to OR operation, if the result is 1, the product is a finished product; if 0, it is the product under process.
The invention has the beneficial effects that: the identification of the content of the finished cigarette products such as whether the goods are in the goods position of the goods in the warehouse, whether the finished cigarette products are wrapped with films or not, whether the finished cigarette products are in the goods position of the warehouse or not, whether the finished cigarette products are wrapped with films or not, whether the cigarette boxes are recycled or not, the number of stack-type cigarette boxes or the number of finished product products in the warehouse or not, and the like are realized. The accuracy and stability are good, and the potential safety hazard problem of manual checking is solved; the checking efficiency is improved, the manpower resource is saved, and unmanned inventory checking is realized; and the real-time image of the cigarette in the checking process is stored to prove the authenticity and effectiveness of the physical checking.
Drawings
FIG. 1 is a schematic representation of several finished cigarette case pattern features of the present invention;
FIG. 2 is a schematic illustration of several material stacks stored in the elevated library of finished cigarette products of the present invention;
figure 3 is a front view and a top pictorial illustration of a standard smoke box 28 of the present invention in a stacked form;
figure 4 is a flow chart illustrating the identification of the number of stacked cartons of the present invention.
Wherein, 1, the position of the pattern of the cigarette case specification; 2. recovering the white label position of the smoke box; 3. the location of the engineering bar code; 4. marking the position of the recovery times; 5. sealing the case with adhesive tape; 6. a tray.
11. Standard smoke box-including various standard grade recycle/normal, hard/soft hardening/short standard smoke box;
12. medium branch smoke box-including recovery/normal, in-process/finished product medium branch smoke box of various product specifications;
13. ramuscule cartons-including recycle/normal, in-process/finished ramuscule cartons of various gauges.
21. 10 pack of standard smoke boxes-including film wrapped/unwrapped;
22. the 20-piece stackers of standard smoke boxes-including film wrapped/unwrapped;
23. a 24 pack of standard smoke boxes-including film wrapped/unwrapped;
24. a 28 pack of standard smoke boxes-including film wrapped/unwrapped;
25. 30 pack of standard smoke boxes-including film wrapped/unwrapped;
26. empty pallet stack type;
27. a 28 piece stack of medium cigarette cases-including film wrapped/unwrapped;
28. a28 piece pack of ramuscule cartons-including film wrapped/unwrapped.
31. A first layer of stack of cigarettes is illustrated in the figures of front view and top view;
32. the first and second layers of cigarette stacks are illustrated in the figures of the main view and the top view;
33. first, second and third layers of cigarette stacks in the figures of the front view and the top view;
34. the first, second, third and fourth layers of cigarette stacks are illustrated in the figures of the principal and top views.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings to facilitate understanding of the skilled person.
A cigarette finished product and cigarette box feature recognition method based on machine vision uses an intelligent camera to recognize whether a goods position of goods in a warehouse has goods, whether a film is wrapped, a goods specification is provided, whether a cigarette box is recycled, the number of the cigarette boxes, the number of products in production/finished products and other content requirements, and compares the content requirements with inventory information to complete warehouse checking work. Its characterized in that, the stock smoke box characteristic that the inventory needs to be used includes: the goods space comprises a cigarette stack (refer to fig. 2), empty goods space inventory information, tray group inventory information, a reflection light spot after the film is wrapped on the cigarette stack, characters, patterns, colors and bar codes on the cigarette box (refer to fig. 1 and 1), white marks attached to the recovered cigarette box (refer to fig. 1 and 2), marks on a fixed position (refer to fig. 1 and 4), a cigarette box stack type main view, a top view (refer to fig. 3), a first cigarette box engineering bar code (refer to fig. 1 and 3) and the like.
The method for identifying whether a real object exists in a cargo space comprises the following steps: and extracting the colors of various cigarette boxes when the goods location has the cigarette stacks to establish a model, searching the color corresponding to the model in a region of interest (ROI), and if the color is found to indicate that the goods location has the cigarette stacks, otherwise, the color is an empty goods location.
The method for identifying whether the cigarette stack is wrapped with the film comprises the following steps: after the cigarette stack is wrapped with the film, the collected image has bright light spots under the angle of illumination light, and according to the characteristic, a gray level histogram analysis algorithm is adopted. And converting the color image into a gray image point by point in the region of interest, calculating and analyzing gray histogram structure data of the region, and determining whether the film is wrapped by adopting contrast in the structure data. Contrast refers to the measurement of different brightness levels between the brightest white and the darkest black of the bright and dark regions in an image, and a larger difference range represents a larger contrast and a smaller difference range represents a smaller contrast.
The effect that whether the empty goods position has a real object or not is greatly influenced by the images of the empty goods position and the tray group, and the identification accuracy is improved by the inventory information of the empty goods position and the inventory information of the tray group.
Method for identifying a quality standard (see fig. 1):
(1) pattern training, the process is as follows: selecting a training pattern, setting a training area and an original point, setting training parameters and a training pattern, and evaluating trained characteristics.
Pattern model training principle: selecting a representative pattern having consistent characteristics; unnecessary features and image noise are reduced; only important features are trained; creating a specific pattern in consideration of the mask; larger patterns provide greater accuracy; the more demarcation points, the higher the accuracy.
(2) The algorithm during the operation: setting parameters during operation, defining a search area, acquiring patterns during operation, operating an intelligent camera algorithm and acquiring a result.
The identification method for whether the smoke box is recovered comprises the following steps: the identification of the recycling bin is marked by pasting a white mark (refer to fig. 1 and 2) on the recycling bin used in the production process and writing numbers (refer to fig. 1 and 4) at fixed positions.
(1) Adopting a pattern matching algorithm, firstly establishing a white mark pattern model, searching a matched target in the image, if the target is 1, identifying the target, and determining a recycling smoke box; if the number is 0, the cigarette is not marked and is a new cigarette box.
(2) For the mode of writing numbers at a fixed position, an OCR character recognition algorithm is adopted, and the number read out at the fixed ROI is the number of times of recycling of the smoke box.
Method for identifying the number of stacked cartons (see fig. 4):
(1) collecting model images of each stack shape and each position, and processing the images;
(2) establishing a main view model (full stack, each layer) and a top view model (each layer and each position);
(3) real-time image processing of the cargo space;
(4) judging whether the primary view is full of stacks, if so, obtaining corresponding stack types and total number according to a primary view full stack model, and if not, entering the next step;
(5) judging whether the main view has four layers or not, if so, counting the number of the fourth layer plus the number of the first three layers according to the top view, and if not, entering the next step;
(6) judging whether the main view has three layers or not, if so, counting the number of the third layer plus the number of the first two layers according to the top view, and if not, entering the next step;
(7) judging whether the main view has two layers or not, if so, counting the number of the second layer plus the number of the previous layer according to the top view, and if not, entering the next step;
(8) and judging whether the main view is one layer or not, if so, counting the number of the first layer according to the top view, and otherwise, judging that the goods position or the tray group is empty.
Take a standard smoke box 28 piece stack as an example (see figure 3):
(1) the first layer contains the placement position shape of each cigarette within 10 cigarettes, the camera shoots an image from a position right above the camera at an angle of 45 degrees, then the image (area division, filtering, balancing and the like) is processed, a model of each cigarette at each position is built, whether each cigarette exists at each position is judged, and then the number of the common cigarettes in the first layer is counted according to the recognition result. The first layer is also modeled on the front side to determine if it is the first layer. (refer to fig. 3, 31).
(2) The second layer contains the placing position shape of each cigarette within 10, the image (area segmentation, filtering, balance and the like) is processed, a overlooking model and a second layer front model of each cigarette are established, the program firstly judges whether the second layer is the overlooking model or not, if the second layer is the overlooking model, the number of the second layer is 10, then the number of the second layer is counted, and the two layers are added to form the total number of the current goods space. (see fig. 3, 32).
(3) The third layer comprises the placing position shapes of all the cigarettes within 4, images (area division, filtering, balance and the like) are processed, an overlooking model and a front model of the third layer of all the cigarettes at all the positions are established, the program firstly judges whether the third layer is available, if the third layer is available, the number of the first layer and the second layer is 10 respectively, the total number of the third layer is 20, and then the number of the third layer is counted, and the three layers are added to form the total number of the current goods space. (refer to fig. 3, 33).
(3) And (3) processing images (area division, filtering, balance and the like) of the placement positions of the cigarettes in the fourth layer, establishing an overlook model and a third layer front model of the cigarettes in each position, judging whether the fourth layer is formed by the program, if the fourth layer is formed, judging that the number of the first layer and the second layer is 10, the number of the third layer is 4, and the total number of the third layer is 24, counting the number of the fourth layer, and adding the four layers to form the total number of the current goods space. (refer to fig. 3, 34).
Method for identifying product in process (see fig. 1): the finished product has the first project code according to the fact that the finished product has no first project code. The identification algorithm of the products in process is to establish a pattern model of the products in process, find the position of the target in the image and read the bar code of the position. Carrying out OR operation on a plurality of results, and if the result is 1, obtaining a finished product; if 0, it is the product under process.
Finally, it is noted that the above-mentioned preferred embodiments illustrate rather than limit the invention, and that, although the invention has been described in detail with reference to the above-mentioned preferred embodiments, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the scope of the invention as defined by the appended claims.
Claims (8)
1. A cigarette finished product cigarette box characteristic identification method based on machine vision uses an intelligent camera to identify whether a goods position of goods in a warehouse is a real object, whether a film is wrapped, a product gauge is adopted, whether a cigarette box is recycled, the number of the cigarette boxes, the content requirements of products in process/finished products and the like are met, and the content requirements are compared with inventory information to complete warehouse checking work; the method is characterized in that: the method comprises a method for identifying whether a goods location has a real object or not; a method for identifying whether the tobacco stack is wrapped with a film; a method for identifying the gauge; a method of identifying whether to recover the smoke box; a method for identifying the number of the stack-type smoke boxes; method for identifying products in process.
2. The machine vision-based finished cigarette box feature identification method according to claim 1, characterized in that: the method requires the use of stock bin features including: the goods position comprises a cigarette stack, empty goods position inventory information, tray group inventory information, a reflection light spot after the cigarette stack is wrapped with a film, characters, patterns, colors and bar codes on the cigarette box, a white mark pasted on the recovered cigarette box, a mark on a fixed position, a cigarette box stack type main view, a top view and a first cigarette box engineering bar code.
3. The machine vision-based finished cigarette case feature identification method of claim 1, characterized by comprising the following steps: the method for identifying whether the goods location has the real goods comprises the following steps: the method comprises the steps of extracting colors of various cigarette boxes when a cigarette stack exists in a goods position, establishing a model, searching colors corresponding to the model in a region of interest (ROI), and if the color is found to indicate that the cigarette stack exists in the goods position, otherwise, the color is an empty goods position, the effect of identifying whether a real object exists in the goods position is greatly influenced by images of the empty goods position and a tray group, and the identification accuracy is improved by empty goods position inventory information and tray group inventory information.
4. The machine vision-based finished cigarette case feature identification method of claim 1, wherein the method comprises the following steps: the method for identifying whether the cigarette stack is wrapped with the film comprises the following steps: after the cigarette stack is wrapped with the film, the collected image has bright light spots under the angle of illumination light, according to the characteristic, a gray level histogram analysis algorithm is adopted, the color image is converted into a gray level image point by point in an interested area, then the gray level histogram structural data of the area is calculated and analyzed, and whether the wrapping film exists or not is determined by adopting the contrast in the structural data.
5. The machine vision-based finished cigarette case feature identification method of claim 1, wherein the method comprises the following steps: the identification method of the gauge comprises the following steps:
pattern training in the step (1), which comprises the following steps: selecting training patterns, setting a training area and an original point, setting training parameters, training the patterns, evaluating trained characteristics, and training a pattern model according to the principle: selecting a representative pattern having consistent characteristics; unnecessary features and image noise are reduced; only important features are trained;
step (2), an algorithm during operation: setting parameters during operation, defining a search area, acquiring patterns during operation, operating an intelligent camera algorithm and acquiring a result.
6. The machine vision-based finished cigarette case feature identification method of claim 1, wherein the method comprises the following steps: the identification method for whether the smoke box is recovered comprises the following steps: the identification of the recycling bin is marked by pasting a white mark on the used recycling bin and writing numbers at a fixed position in the production process;
step (1) adopting a pattern matching algorithm, firstly establishing a white mark pattern model, searching a matched target in an image, if the matched target is 1, identifying the target, and determining a recycling smoke box; if the number is 0, the cigarette is not marked, and the cigarette is a new cigarette box;
and (2) reading in a fixed ROI by adopting an OCR character recognition algorithm for a mode of writing numbers in a fixed position, wherein the number read out is the recycling number of the smoke box.
7. The machine vision-based finished cigarette case feature identification method of claim 1, wherein the method comprises the following steps: the method for identifying the number of the stack type cigarette boxes comprises the following steps:
collecting model images of each stack shape and each position, and processing the images;
step (2) establishing a main view model (full stack, each layer) and a overlook model (each layer and each position);
step (3) real-time image processing of the cargo space;
step (4) judging whether the primary view is full of stacks, if so, obtaining corresponding stack types and total number according to a primary view full stack model, and if not, entering the next step;
step (5) judging whether the main view has four layers, if so, counting the number of the fourth layer plus the number of the first three layers according to the main view, and if not, entering the next step;
step (6) judging whether the main view has three layers or not, if so, counting the number of the third layer plus the number of the first two layers according to the main view, and if not, entering the next step;
step (7) judging whether the main view has two layers, if so, counting the number of the second layer plus the number of the previous layer according to the top view, and if not, entering the next step;
and (8) judging whether the main view is a layer or not, counting the number of the first layer according to the top view if the main view is the layer, and judging the goods position or the tray group to be empty if the main view is the layer.
8. The machine vision-based finished cigarette case feature identification method of claim 1, wherein the method comprises the following steps: the identification method of the products in process/finished products comprises the following steps: according to the fact that no one project code exists in a product, a finished product has the one project code, a product in process/finished product identification algorithm is used, a pattern model of the product in process is established firstly, the position of a target is found in an image, then a bar code of the position is read, a plurality of results are subjected to OR operation, and if the number of the results is 1, the product is obtained; if 0, it is the product under process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210194180.6A CN114743082A (en) | 2022-03-01 | 2022-03-01 | Cigarette finished product cigarette box characteristic identification method based on machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210194180.6A CN114743082A (en) | 2022-03-01 | 2022-03-01 | Cigarette finished product cigarette box characteristic identification method based on machine vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114743082A true CN114743082A (en) | 2022-07-12 |
Family
ID=82274627
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210194180.6A Pending CN114743082A (en) | 2022-03-01 | 2022-03-01 | Cigarette finished product cigarette box characteristic identification method based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114743082A (en) |
-
2022
- 2022-03-01 CN CN202210194180.6A patent/CN114743082A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115272341B (en) | Packaging machine defect product detection method based on machine vision | |
| CN107617573B (en) | Logistics code identification and sorting method based on multitask deep learning | |
| CN109658412B (en) | Rapid packaging box identification and segmentation method for unstacking and sorting | |
| US7356162B2 (en) | Method for sorting postal items in a plurality of sorting passes | |
| CN111310645B (en) | Method, device, equipment and storage medium for warning overflow bin of goods accumulation | |
| Hu et al. | Defect identification method for poplar veneer based on progressive growing generated adversarial network and MASK R-CNN Model | |
| CN113821674B (en) | Intelligent cargo supervision method and system based on twin neural network | |
| CN111126393A (en) | Vehicle appearance refitting judgment method and device, computer equipment and storage medium | |
| CN112116582A (en) | Cigarette detection and identification method under stock or display scene | |
| CN113379720A (en) | Tea cake anti-counterfeiting method based on tea cake image feature code | |
| CN109255414A (en) | A kind of colour barcode made an inventory for books, books recognition methods, electronic equipment and storage medium | |
| CN116052222A (en) | Cattle face recognition method for naturally collecting cattle face image | |
| CN114511718A (en) | Intelligent management method and system for materials for building construction | |
| CN111597857A (en) | Logistics package detection method, device and equipment and readable storage medium | |
| CN114332622A (en) | Label detection method based on machine vision | |
| CN113392747A (en) | Goods packing box identification method and system for stereoscopic warehouse | |
| CN114743082A (en) | Cigarette finished product cigarette box characteristic identification method based on machine vision | |
| CN117152731A (en) | Method and system for identifying key information of label of warehouse-in material box of cigarette factory | |
| CN115619320A (en) | Product external packing informatization error-preventing system | |
| CN114800533B (en) | Sorting control method and system for industrial robot | |
| CN115082832A (en) | Information identification method, device and storage medium | |
| CN115880477A (en) | Apple detection positioning method and system based on deep convolutional neural network | |
| CN114996500B (en) | Trademark graph retrieval method | |
| CN109409453A (en) | A kind of cartoning sealing machine lacks a traceability system and its means of proof | |
| Castano-Amoros et al. | MOSPPA: monitoring system for palletised packaging recognition and tracking |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |