CN116091825A - Nixie tube content identification method and device, electronic equipment and storage medium - Google Patents
Nixie tube content identification method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116091825A CN116091825A CN202310019448.7A CN202310019448A CN116091825A CN 116091825 A CN116091825 A CN 116091825A CN 202310019448 A CN202310019448 A CN 202310019448A CN 116091825 A CN116091825 A CN 116091825A
- Authority
- CN
- China
- Prior art keywords
- nixie tube
- brightness
- identification
- nixie
- display section
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/60—Extraction of image or video features relating to illumination properties, e.g. using a reflectance or lighting model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/02—Recognising information on displays, dials, clocks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of digital identification, in particular to a nixie tube content identification method, a nixie tube content identification device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring a continuous video image sequence of the nixie tube, and acquiring position information of a character display section of the nixie tube according to a preset frame number by using a template matching algorithm; recognizing the bright and dark states of the character display section of the nixie tube by using a Support Vector Machine (SVM); acquiring a brightness sequence unit of a nixie tube character display section under a preset frame number according to a preset nixie tube arrangement mode; classifying the brightness sequence units by using a deep learning method and outputting a recognition result; the device comprises: the digital tube signal processing device comprises a digital tube and a PC end connected with the digital tube signal, wherein the digital tube and the PC end are configured into an unsynchronized state, and the PC end comprises: the device comprises an acquisition module, a position identification unit, a brightness identification unit and a training module. The method aims to solve the problem of false identification caused by uneven brightness and incomplete brightness in the process of identifying the nixie tube content.
Description
Technical Field
The invention relates to the technical field of digital identification, in particular to a nixie tube content identification method, a nixie tube content identification device, electronic equipment and a storage medium.
Background
The nixie tube is one of LED display screens, is widely applied to meters such as voltmeters, ammeter, thermometer and the like in a transformer substation due to low price and convenient use, and displays numerical information of related meters through the nixie tube; in the existing nixie tube LED display screen, the display units of the LED display screen are displayed asynchronously due to the driving mode principle; for example, an auxiliary monitoring system of a transformer substation needs to carry out all-weather inspection on equipment meters of the transformer substation, so that the accurate perception of the states and parameters of the equipment meters is realized, the manual inspection intensity is reduced, and finally unmanned inspection is realized; the cameras installed in the existing auxiliary monitoring system are usually commercial cameras, the control parameters are few, no physical link linkage exists between the cameras and the LED display screen, or the cameras and the LED display screen are mutually independent units; when the two are operated in an asynchronous state, the problem that images acquired by a camera are always in a incomplete or incomplete state or display brightness difference is caused, and the images cannot be identified, so that the identification accuracy of the content of the LED display screen is greatly reduced;
the invention patent nixie tube dynamic display identification method and system with the publication number of CN108182400A is characterized in that the dynamic display period of the nixie tube is acquired, the dynamic scanning period of the nixie tube is set as the exposure time of a camera through a PC end, so that the complete acquisition of the nixie tube content is realized, and further, the identification is realized, but the mode has higher requirements on the camera and the system, the camera is required to be controllable, the LED display screen is required to be synchronously linked with the system, and the scheme can solve the problem of image defect acquisition, but has a certain difficulty in system deployment.
Therefore, we propose a nixie tube content identification method, a nixie tube content identification device, an electronic device and a storage medium.
Disclosure of Invention
The invention aims to provide a nixie tube content identification method, a nixie tube content identification device, electronic equipment and a storage medium, which are used for solving the problem of false identification caused by uneven brightness and incomplete brightness in the nixie tube content identification process.
The technical scheme of the first aspect of the invention provides a nixie tube content identification method, which comprises the following steps:
acquiring a continuous video image sequence of the nixie tube, and acquiring position information of a character display section of the nixie tube according to a preset frame number by using a template matching algorithm;
recognizing the bright and dark states of the character display section of the nixie tube by using a Support Vector Machine (SVM);
acquiring a brightness sequence unit of a nixie tube character display section under a preset frame number according to a preset nixie tube arrangement mode;
and classifying the brightness sequence units by using a deep learning method and outputting the recognition result.
Further, the template matching algorithm specifically includes: the method for matching and identifying the position information of the display segment by utilizing the corner matching method comprises the following specific steps:
acquiring original image information of a nixie tube and target image information in a video image sequence, and extracting key feature points by using a feature point extraction algorithm;
positioning the key feature points, extracting feature vectors, and comparing to obtain mutually matched key feature points;
and eliminating noise by utilizing a RANSAC algorithm aiming at the key feature points matched with each other so as to acquire the specific position information of the display segment.
Further, the key feature points include at least corner points, edge points, two points of a dark area, and a dark point of a bright area.
Further, the preset nixie tube arrangement mode specifically includes: and taking the sequence of refreshing display segments of the nixie tube as an arrangement rule of the brightness sequence units.
Further, the method for performing the deep learning on the brightness sequence units and outputting the recognition result specifically includes:
manually labeling the brightness sequence units and obtaining a labeling data set;
training based on the labeling data set to obtain a classification model;
and identifying the brightness sequence unit by using the classification model and outputting an identification result.
Further, the training based on the labeling data set to obtain the classification model specifically includes:
and training the labeling data set by utilizing the Resnet-50 convolutional neural network model to obtain a classification model.
The technical scheme of the second aspect of the invention provides a nixie tube content recognition device, which comprises a nixie tube and a PC end connected with the nixie tube in a signal manner, wherein the nixie tube and the PC end are configured into an unsynchronized state, and the PC end comprises:
the acquisition module is configured to acquire a continuous video image sequence of the nixie tube;
the position identification unit is configured to acquire position information of a character display section of the nixie tube according to a preset frame number by using a template matching algorithm;
the brightness recognition unit is configured to obtain a brightness sequence unit of a nixie tube character display section under a preset frame number according to a preset nixie tube arrangement mode;
and the training module is configured to classify the brightness sequence units by using a deep learning method and output the recognition result.
A technical solution of a third aspect of the present invention provides an electronic device, including: a processor and a memory communicatively coupled to the processor; the memory stores instructions executable by the processor, so that the processor can execute steps of the nixie tube content identification method according to any one of the technical schemes of the first aspect of the invention.
A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon a program for implementing a nixie tube content identification method, the program for implementing the nixie tube content identification method being executed by a processor to implement the steps of the content identification method according to any one of the first aspect of the present invention.
The beneficial effects of the invention include:
1. according to the nixie tube content identification method provided by the invention, the bright and dark states of the character display section of the nixie tube are identified through the support vector machine SVM multiframe, so that the problem of false identification caused by uneven brightness and incomplete brightness of the nixie tube is solved; by acquiring the brightness sequence units, classifying the brightness sequence units by using a deep learning method and outputting the recognition result, the content recognition of the nixie tube and the image acquisition device in an unsynchronized state is realized, the requirement on the performance of the image acquisition device is reduced, the deployment is simple, and the accuracy of recognizing the content of the nixie tube is effectively improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments of the present invention will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and should not be considered as limiting the scope, and other related drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a nixie tube content identification method provided by an embodiment of the present invention;
FIG. 2 is a flowchart of a template matching algorithm provided in an embodiment of the present invention;
fig. 3 is a schematic structural diagram of a nixie tube content recognition device according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a bottleneck residual block according to an embodiment of the present invention;
fig. 5 is a schematic diagram of a network structure of a Resnet-50 convolutional neural network model according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described below with reference to the accompanying drawings in the embodiments of the present invention.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures. Meanwhile, in the description of the present invention, the terms "first", "second", and the like are used only to distinguish the description, and are not to be construed as indicating or implying relative importance.
Referring to fig. 1, the technical solution of the first aspect of the present invention provides a nixie tube content identification method, including:
step S1: acquiring a continuous video image sequence of the nixie tube, and acquiring position information of a character display section of the nixie tube according to a preset frame number by using a template matching algorithm;
in step S1, in order to realize stable identification of the nixie tube, on the premise that parameters of an unknown image acquisition device or parameters of the image acquisition device are uncontrollable, continuous video images of the nixie tube in 1S are acquired and converted into a video image sequence; the preset frame number can be a single frame or a plurality of frames;
preferably, the direct of the nixie tube can be marked by a manual marking mode by acquiring the length, width and height information of the nixie tube, and then the position information of the corresponding display section of the nixie tube in the video image sequence is matched by a mapping method;
step S2: recognizing the bright and dark states of the character display section of the nixie tube by using a Support Vector Machine (SVM);
the method comprises the steps of utilizing a Support Vector Machine (SVM) to identify, through multi-frame identification of a video image sequence, avoiding the problem of false identification caused by uneven darkness of a nixie tube character display Duan Liang, specifically, marking the bright and dark state of each character display section by adopting a ResNet-50 neural network algorithm, and using output 0 or 1 to represent, wherein 0 represents darkness and 1 represents brightness;
step S3: acquiring a brightness sequence unit of a nixie tube character display section under a preset frame number according to a preset nixie tube arrangement mode;
step S3 is based on the bright and dark states of the character display sections of the nixie tubes in step S2, and a brightness sequence unit is formed according to a preset nixie tube arrangement mode; for example, "01000001" is a complete string of luminance sequence units; the first seven bits of the brightness sequence unit represent the bright and dark states of seven character display sections of the nixie tube; the eighth bit of the brightness sequence unit represents the bright and dark state of the decimal point of the nixie tube;
step S4: classifying the brightness sequence units by using a deep learning method and outputting a recognition result;
the output identification result is a number actually displayed by a nixie tube, wherein step S4 needs to label the brightness sequence unit based on semantic information in the continuous video image sequence, such as "1.", and the like, and the semantic information comprises information of a scene, a number, a letter, and the like to be identified;
according to the nixie tube content identification method provided by the embodiment, the bright and dark states of the character display section of the nixie tube are identified through the support vector machine SVM multiframe, so that the problem of false identification caused by uneven brightness and incomplete brightness of the nixie tube is solved; the brightness sequence units are obtained, the deep learning method is utilized to classify the brightness sequence units and output the identification result, so that the content identification of the nixie tube and the image acquisition device in the unsynchronized state is realized; in practical application, the requirements on image acquisition device equipment such as a camera are reduced, the content identification of the nixie tube is improved from an algorithm layer as a whole, the accuracy of the content identification of the nixie tube is effectively improved through multi-frame identification and a deep learning method, and meanwhile, the deployment of an identification system is easy;
referring to fig. 2, the template matching algorithm preferably includes: the method for matching and identifying the position information of the display segment by utilizing the corner matching method comprises the following specific steps:
acquiring original image information of a nixie tube and target image information in a video image sequence, and extracting key feature points by using feature point extraction algorithms such as SIFT, harriiH corner points and the like;
specifically, the SIFT algorithm is utilized to identify the image positions on all scale spaces in the original image information through a Gaussian differential function, and potential interest points with scale and rotation invariance are identified;
positioning the key feature points, extracting feature vectors, and comparing to obtain mutually matched key feature points;
the key feature points are positioned at the positions of each preset key feature point, and the positions of the key feature points can be determined by fitting a model;
eliminating noise by utilizing a RANSAC algorithm aiming at key feature points matched with each other so as to acquire specific position information of a display segment;
in the embodiment, since the relative positions of the acquisition equipment and the nixie tube are not fixed in actual work, the acquired images of the nixie tube have size difference, so that a template matching algorithm with scale invariance is selected;
preferably, the key feature points at least comprise corner points, edge points, two points of a dark area and a dark point of a bright area; in this embodiment, the key feature points are points that do not disappear due to factors such as illumination, size, rotation, and the like;
preferably, the preset nixie tube arrangement mode specifically includes: and taking the sequence of refreshing display segments of the nixie tube as an arrangement rule of the brightness sequence units.
Preferably, the performing the deep learning method on the luminance sequence units and outputting the recognition result specifically includes:
manually labeling the brightness sequence units and obtaining a labeling data set;
training based on the labeling data set to obtain a classification model;
and identifying the brightness sequence unit by using the classification model and outputting an identification result.
Further, the training based on the labeling data set to obtain the classification model specifically includes:
training a labeling data set by utilizing a Resnet-50 convolutional neural network model to obtain a classification model;
wherein, the Resnet-50 convolutional neural network model comprises 49 convolutional layers and 1 fully-connected layer; as shown in fig. 4, fig. 4 is a bottleneck residual block, which is a basic module element of Resnet-50, and the number of parameters in the model can be greatly reduced through dimension up and down operations; as shown in fig. 5, fig. 5 is a network structure diagram of Resnet-50, the input of the network is 224×224×3, the output is 7×7×2048 after the convolution calculation of the first five parts, the pooling layer will convert it into a feature vector, and finally the classifier will calculate the feature vector and output the class probability;
referring to fig. 3, the technical solution of the second aspect of the present invention provides a nixie tube content identification device, including a nixie tube and a PC end connected with the nixie tube by signals, wherein the nixie tube and the PC end are configured in an unsynchronized state, and the PC end includes:
the acquisition module is configured to acquire a continuous video image sequence of the nixie tube;
the position identification unit is configured to acquire position information of a character display section of the nixie tube according to a preset frame number by using a template matching algorithm;
the brightness recognition unit is configured to obtain a brightness sequence unit of a nixie tube character display section under a preset frame number according to a preset nixie tube arrangement mode;
and the training module is configured to classify the brightness sequence units by using a deep learning method and output the recognition result.
A technical solution of a third aspect of the present invention provides an electronic device, including: a processor and a memory communicatively coupled to the processor; the memory stores instructions executable by the processor, so that the processor can execute steps of the nixie tube content identification method according to any one of the technical schemes of the first aspect of the invention.
A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon a program for implementing a nixie tube content identification method, the program for implementing the nixie tube content identification method being executed by a processor to implement the steps of the content identification method according to any one of the first aspect of the present invention.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.
Claims (9)
1. The nixie tube content identification method is characterized by comprising the following steps of:
acquiring a continuous video image sequence of the nixie tube, and acquiring position information of a character display section of the nixie tube according to a preset frame number by using a template matching algorithm;
recognizing the bright and dark states of the character display section of the nixie tube by using a Support Vector Machine (SVM);
acquiring a brightness sequence unit of a nixie tube character display section under a preset frame number according to a preset nixie tube arrangement mode;
and classifying the brightness sequence units by using a deep learning method and outputting the recognition result.
2. The method of claim 1, wherein the template matching algorithm specifically comprises: the method for matching and identifying the position information of the display segment by utilizing the corner matching method comprises the following specific steps:
acquiring original image information of a nixie tube and target image information in a video image sequence, and extracting key feature points by using a feature point extraction algorithm;
positioning the key feature points, extracting feature vectors, and comparing to obtain mutually matched key feature points;
and eliminating noise by utilizing a RANSAC algorithm aiming at the key feature points matched with each other so as to acquire the specific position information of the display segment.
3. The identification method according to claim 2, wherein the key feature points include at least corner points, edge points, two points of a dark area, and a dark point of a bright area.
4. The identification method according to claim 1, wherein the preset nixie tube arrangement mode specifically includes: and taking the sequence of refreshing display segments of the nixie tube as an arrangement rule of the brightness sequence units.
5. The recognition method according to claim 1, wherein the performing of the respective and output of the recognition result on the luminance sequence units by the deep learning method specifically comprises:
manually labeling the brightness sequence units and obtaining a labeling data set;
training based on the labeling data set to obtain a classification model;
and identifying the brightness sequence unit by using the classification model and outputting an identification result.
6. The method of claim 5, wherein training based on the labeled dataset to obtain the classification model specifically comprises:
and training the labeling data set by utilizing the Resnet-50 convolutional neural network model to obtain a classification model.
7. The utility model provides a nixie tube content recognition device which characterized in that includes the nixie tube and with the PC end of nixie tube signal connection, dispose to asynchronous state between nixie tube and the PC end, the PC end includes:
the acquisition module is configured to acquire a continuous video image sequence of the nixie tube;
the position identification unit is configured to acquire position information of a character display section of the nixie tube according to a preset frame number by using a template matching algorithm;
the brightness recognition unit is configured to obtain a brightness sequence unit of a nixie tube character display section under a preset frame number according to a preset nixie tube arrangement mode;
and the training module is configured to classify the brightness sequence units by using a deep learning method and output the recognition result.
8. An electronic device, the electronic device comprising: a processor and a memory communicatively coupled to the processor; wherein the memory stores instructions executable by the processor to enable the processor to perform the steps of the nixie tube content identification method of any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a program for realizing a nixie tube content recognition method, the program for realizing the nixie tube content recognition method being executed by a processor to realize the steps of the content recognition method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310019448.7A CN116091825A (en) | 2023-01-06 | 2023-01-06 | Nixie tube content identification method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310019448.7A CN116091825A (en) | 2023-01-06 | 2023-01-06 | Nixie tube content identification method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116091825A true CN116091825A (en) | 2023-05-09 |
Family
ID=86203933
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310019448.7A Pending CN116091825A (en) | 2023-01-06 | 2023-01-06 | Nixie tube content identification method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116091825A (en) |
-
2023
- 2023-01-06 CN CN202310019448.7A patent/CN116091825A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109086756B (en) | Text detection analysis method, device and equipment based on deep neural network | |
| Laroca et al. | Convolutional neural networks for automatic meter reading | |
| CN107016387B (en) | Method and device for identifying label | |
| US10445602B2 (en) | Apparatus and method for recognizing traffic signs | |
| CN110555420B (en) | Fusion model network and method based on pedestrian regional feature extraction and re-identification | |
| CN111700608A (en) | Multi-classification method and device for electrocardiosignals | |
| CN111340796B (en) | Defect detection method and device, electronic equipment and storage medium | |
| CN114155527A (en) | Scene text recognition method and device | |
| CN109753962B (en) | Method for processing text region in natural scene image based on hybrid network | |
| CN113989794B (en) | License plate detection and recognition method | |
| CN109508714B (en) | Low-cost multi-channel real-time digital instrument panel visual identification method and system | |
| CN113409355A (en) | Moving target identification system and method based on FPGA | |
| CN104182752A (en) | Intelligent monitoring method of outdoor advertising board | |
| CN114972246A (en) | Die-cutting product surface defect detection method based on deep learning | |
| US10055668B2 (en) | Method for the optical detection of symbols | |
| CN108960005B (en) | Method and system for establishing and displaying object visual label in intelligent visual Internet of things | |
| CN116580326A (en) | Aviation environment safety risk prevention and control detection and early warning system | |
| CN116091825A (en) | Nixie tube content identification method and device, electronic equipment and storage medium | |
| CN114627435B (en) | Intelligent light adjusting method, device, equipment and medium based on image recognition | |
| Shi et al. | An invoice recognition system using deep learning | |
| CN115843375A (en) | Logo labeling method and device, logo detection model updating method and system and storage medium | |
| CN112464946A (en) | Garbage image collection method and system based on big data | |
| CN111612094A (en) | Speed false detection and correction method, equipment and computer readable storage medium | |
| CN104866825B (en) | A kind of sign language video frame sequence classification method based on Hu square | |
| CN112163800B (en) | Management method and device for visual inspection tool |
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 |