CN111401139A - Method for obtaining position of underground mine equipment based on character image intelligent identification - Google Patents
Method for obtaining position of underground mine equipment based on character image intelligent identification Download PDFInfo
- Publication number
- CN111401139A CN111401139A CN202010114364.8A CN202010114364A CN111401139A CN 111401139 A CN111401139 A CN 111401139A CN 202010114364 A CN202010114364 A CN 202010114364A CN 111401139 A CN111401139 A CN 111401139A
- Authority
- CN
- China
- Prior art keywords
- image
- character
- convolution
- network
- images
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 17
- 230000003137 locomotive effect Effects 0.000 claims abstract description 21
- 238000004519 manufacturing process Methods 0.000 claims abstract description 21
- 238000005070 sampling Methods 0.000 claims abstract description 7
- 238000001514 detection method Methods 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 27
- 230000004913 activation Effects 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 6
- 238000013135 deep learning Methods 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000005065 mining Methods 0.000 abstract description 8
- 238000005516 engineering process Methods 0.000 description 6
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 239000000428 dust Substances 0.000 description 2
- 238000012937 correction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000005272 metallurgy Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
Abstract
The invention provides a method for intelligently identifying and obtaining the position of underground equipment of a mine based on character images, which is characterized in that a plurality of character boards are installed beside an underground rail at intervals, a plurality of characters are marked on each character board, and the position of the rail corresponding to the serial number of each character is recorded in a database of a production scheduling center; installing image acquisition equipment on an unmanned locomotive under a mine, and acquiring all character images marked on corresponding character boards during running; the collected images are delivered to a U-Net network for detection, after the images with 8 characters are detected, the images are divided into 8 non-overlapping sub-images, each sub-image should contain 1 character, classification and identification are carried out after convolution operation and down-sampling operation, corresponding character values and credibility values are obtained, the character values corresponding to the positions with the maximum credibility values are transmitted to a production scheduling center in real time through a wireless network, accurate positioning can be carried out on the unmanned locomotive underground, and the requirements of industrialization, automatic mining and transportation production are met.
Description
Technical Field
The invention relates to a method for obtaining the position of underground equipment of a mine, in particular to a method for obtaining the position of underground equipment of the mine based on character image intelligent identification, and belongs to the technical field of character image identification.
Background
The mining industry is an upstream industry of the metallurgical industry, provides main raw materials for metallurgy, is a capital-intensive, resource-intensive, technology-intensive industry, and is a high-energy-consumption industry. With the continuous expansion of the production scale of enterprises, the production safety in the operation process is increasingly emphasized, various measures are actively taken to innovate the safety management mode and method of the enterprises, and the safety management level of the enterprises is continuously improved. The underground operation of the mine is unmanned, which is an important safety measure for mining production, the unmanned underground mining operation is completely automatically completed by automatic equipment, and because the underground distance reaches hundreds of kilometers and the number of fork openings is very large, the mining automatic equipment needs to be positioned, the position of the automatic mining equipment under the mine needs to be mastered, the automatic mining equipment is convenient for a production scheduling center monitoring equipment, and the equipment is scheduled to perform production operation scientifically in real time according to the production condition.
The unmanned locomotive is an electric locomotive for transporting ores under a mine, the electric locomotive runs on an underground tunnel rail, the length of the rail reaches hundreds of kilometers, and after the electric locomotive is transformed into the unmanned electric locomotive, the electric locomotive automatically runs on the rail, so that the position of the electric locomotive on the rail needs to be known in real time. There are many ways of positioning unmanned vehicles, such as active signal technologies like RFID, but RFID devices need to be installed under the well of hundreds of kilometers, which is too high in cost, and the wireless signals are greatly interfered under the well, which is not beneficial to actual production. With the progress of artificial intelligence technology, it is possible to solve the positioning problem by using character recognition, namely, installing 8 character length plates at every 20 meters beside a rail, wherein the size of each plate is similar to that of a vehicle license plate, when a vehicle passes through the character plates, the image acquisition equipment on the unmanned vehicle detects the corresponding character, then segmenting and classifying recognition are carried out, the recognition result is transmitted to a production scheduling center through a wireless network, and the scheduling center can master the specific position of the unmanned vehicle underground according to the character number. For underground character image recognition of a mine, a general character recognition algorithm cannot be used, and the traditional digital image processing technology cannot be well adapted to various complex environments, for example, the imaging effect can be influenced by low underground light intensity, and the character surface characteristics can be influenced by serious underground dust. Therefore, there is a need for improvements in the prior art.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a method for carrying out intelligent image identification based on image data acquired by image acquisition equipment and combining a deep learning technology to finally obtain the specific position of an underground locomotive.
The invention is completed by the following technical scheme: a method for obtaining the position of underground equipment of a mine based on character image intelligent identification is characterized by comprising the following steps:
1) installing a plurality of character boards beside a rail under a mine at intervals, marking a plurality of characters on each character board, and recording the rail position corresponding to each character number in a database of a production scheduling center;
2) the image acquisition equipment is arranged on an unmanned locomotive under a mine, and all character images marked on corresponding character boards are acquired during running;
3) reading the acquired character image data by using a video Capture class in an open source image processing library OpenCV, wherein the frame rate is 15 frames per second, the pixel format is an RGB three-channel image, and the original pixel is 1024900;
4) carrying out conventional zooming and filtering pretreatment on each frame of image data, wherein the zooming treatment is to zoom or enlarge the image size so as to reduce the processing amount of deep learning network model data and accelerate the segmentation and identification speed of each frame of image, and the L anczos algorithm is used for compressing the image into 800600 pixels in the form of RGB color image;
5) the preprocessed 800600-pixel image is delivered to a U-Net network for detection, after 8-character images in the image are detected, the image is divided into 8 non-overlapping sub-images, and each sub-image contains 1 character;
the U-Net network is an Encoder-Decoder network structure, wherein: the method comprises the following steps that an Encoder network is used for convolution operation, a Decode network is used for up-sampling operation, the Encoder network is of a five-layer convolution network structure, the convolution kernel of each layer of convolution network structure is 55, padding is 0, striping is 1, the Decode network is of a five-layer convolution network structure, the convolution kernel of each layer of convolution network is 11, and the step length is 1;
the convolution operation and the downsampling operation are performed using the following algorithm:
5-1) convolution operation as follows: processing 800600 pixel image into 780580 pixel by five-layer convolution operation, and pooling with step size 2 to 390290 pixel by 22 convolution kernel; repeating the operation for three times to obtain a pixel with an image of 6045; the convolution operation is formulated as follows:
where X is the image data, i and j are the image sizes, 800 and 600 respectively, W is the convolution kernel, m and n are the convolution kernel sizes, here 5 and 5 respectively, and s (i, j) is the new image data after the convolution operation;
after each convolution operation, performing nonlinear calculation by using an activation function, wherein the activation function of the whole network uses a Maxout activation function, and the activation function formula is as follows:
wherein z isij＝xTw...ij+bij+ c, wherein xTIs the network neuron number, WijIs convolution kernel value, i and j are coordinate positions in the convolution kernel, k is the number of channels of the image, the image is an RGB color image, k is 3, bij is a constant corresponding to each neuron, c is an empirical constant after activation calculation, the initial value 0, j is a subscript and is max Zij；
Because the convolution operation is linear operation, the nonlinear processing is carried out by using a loss function, the loss function uses pixel-wise softmax, the output corresponding to the pixel is independently made into softmax, and the formula is as follows:
in the formula, x is a pixel position on a two-dimensional plane, a is a learning coefficient, an initial value is 1, w (x) is a weight term in cross entropy, pl (x) represents the output probability of x on a channel where a real label is located, c is a constant term, and the initial value is 0;
5-2) the downsampling operation is as follows: sending the image with 6045 pixels obtained by convolution operation into a Decoder network, increasing the length and width of the image into 12090 pixels by two times after upsampling operation, repeating the two times of the sampling operation on the Decoder network to obtain 480360 pixels, performing convolution operation by using 55 convolution kernels to obtain 420300 pixels of the image, performing convolution operation by using 51 convolution kernels to obtain 400300 pixels of the image, and performing upsampling operation once to recover 800600 pixels which are the same as the original image;
5-3) carrying out full connection operation on the recovered 800600 acorn image to obtain the specific position of a single character in the original image, and expressing the position by using the upper left corner coordinate and the lower right corner coordinate of the sub-image;
the full join operation is as follows: the positions of 8 character images are mainly output by full connection, each image coordinate consists of 4 values in the upper left corner and the lower right corner, and the 8 images have 32 output values; the image is a two-dimensional array of 800600, which is converted into a one-dimensional array by rows with length of 480000, then 32 one-dimensional array parameters with length of 48000 are multiplied and summed with the image one-dimensional array pixel value, and an intercept parameter is added, so that the obtained 32 values are the coordinate positions of 8 characters;
full connection formula: x is the number ofi＝anm*wi+ci
In the formula xiIs 32 coordinate values, i takes a value from 1 to 32, anmIs a one-dimensional array of images, wiIs a one-dimensional array parameter, length n m, ciIs an intercept parameter, wiAnd ciAre both learnable parameters, n and m are the length and width of the source image, respectively, i.e. n is 800 and m is 600;
6) classification and identification: respectively subjecting the sub-images which are segmented in the step 5) and contain single characters to convolutional neural network to sequential classification and recognition, wherein each sub-image is recognized once to obtain a character value and a credible value, and 8 classification and recognition are carried out for 8 times to obtain 8 character values and 8 corresponding credible values, and each credible value is larger than 90%;
the convolutional neural network structure has 8 layers, wherein: the 1-3 layer network uses 9 types of convolution kernels to extract 9 characteristics, each type of convolution kernel is 33, the 4-6 layer network uses 12 types of convolution kernels, each convolution kernel is 33, the 7 th layer of convolution kernel uses 1024 types of convolution kernels, each convolution kernel is 11, the 8 th layer is a full-connection layer, 62 credible values are output, and the character corresponding to the position with the maximum credible value is the recognized character value;
after each convolution operation, carrying out nonlinear operation by using an activation function, wherein the activation function uses an exponential linear unit E L U function;
7) the obtained 8 character values are transmitted to a production scheduling center in real time through a wireless network, the character numbers are used as query conditions, the specific positions are looked up in a database system, the underground positions of the unmanned locomotives can be determined, the underground positioning of the unmanned locomotives in the mine is achieved, meanwhile, image data with low probability values of recognition results are automatically stored, the stored images are trained in a targeted mode every month, the training rate is 99.89%, the network model on production is updated, and the purpose of continuous learning is achieved.
The invention has the following beneficial effects:
the invention solves the technical problems of large error, high cost and the like caused by unclear images due to insufficient light receiving and large dust amount of underground automatic equipment, can greatly improve the underground character image recognition rate to 99.89 percent by the invention, combines the recognition error correction technology, and can reach 100 percent by the recognition rate, thereby accurately positioning the underground unmanned automatic locomotive, meeting the requirements of industrialization, automatic mining and transportation production, saving investment and being widely applied to underground operation of mines and other industries of industrial production and logistics transportation.
Detailed description of the preferred embodiments
The following detailed description will describe embodiments of the invention, which are exemplary only and not to be construed as limiting the invention.
Examples
The image acquisition equipment of the embodiment uses an Nvidia TX2 artificial intelligence edge calculation device, the size of the equipment is 50 mm x 87 mm, the size is small, the power consumption is only 7.5 watts, the equipment is installed on a mobile unmanned locomotive, the space occupied by installation is small, the power consumption is low, and the method which is not described in detail is a conventional technology.
The method comprises the following specific steps:
1) the method is characterized in that a character plate is installed at intervals of 20 meters beside a rail under a mine, the size of the character plate is similar to that of a license plate of an automobile, 8 characters are printed on each license plate, and the character plate is installed at each intersection. Recording the position of each numbered rail in a database system, wherein the database system is arranged in a production scheduling center;
2) the method comprises the following steps of (1) installing image acquisition equipment on an unmanned locomotive under a mine to carry out image acquisition work, and acquiring 6 thousand underground character images, wherein 5 thousand underground character images are used for training a network model, and 1 thousand underground character images are used for testing the network model;
3) reading video image data of a camera by using a video Capture type in an open source image processing library OpenCV, wherein the frame rate is 15 frames per second, the pixel format is an RGB three-channel image, and the original pixel is 1024900;
4) carrying out conventional zooming and filtering pretreatment on each frame of image data, wherein the zooming treatment is to zoom or enlarge the image size so as to reduce the processing amount of deep learning network model data and accelerate the segmentation and identification speed of each frame of image, and the L anczos algorithm is used for compressing the image into 800600 pixels in the form of RGB color image;
5) the preprocessed 800600-pixel image is delivered to a U-Net network for detection, after 8-character images in the image are detected, the image is divided into 8 non-overlapping sub-images, and each sub-image contains 1 character;
the U-Net network is an Encoder-Decoder network structure, wherein: the method comprises the following steps that an Encoder network is used for convolution operation, a Decode network is used for up-sampling operation, the Encoder network is of a five-layer convolution network structure, the convolution kernel of each layer of convolution network structure is 55, padding is 0, striping is 1, the Decode network is of a five-layer convolution network structure, the convolution kernel of each layer of convolution network is 11, and the step length is 1;
the convolution operation and the downsampling operation are performed using the following algorithm:
5-1) convolution operation as follows: processing 800600 pixel image into 780580 pixel by five-layer convolution operation, and pooling with step size 2 to 390290 pixel by 22 convolution kernel; repeating the operation for three times to obtain a pixel with an image of 6045; the convolution operation is formulated as follows:
where X is the image data, i and j are the image sizes, 800 and 600 respectively, W is the convolution kernel, m and n are the convolution kernel sizes, here 5 and 5 respectively, and s (i, j) is the new image data after the convolution operation;
after each convolution operation, performing nonlinear calculation by using an activation function, wherein the activation function of the whole network uses a Maxout activation function, and the activation function formula is as follows:
wherein z isij＝xTw...ij+bij+ c, wherein xTIs the network neuron number, WijIs convolution kernel value, i and j are coordinate positions in the convolution kernel, k is the number of channels of the image, the image is an RGB color image, k is 3, bij is a constant corresponding to each neuron, c is an empirical constant after activation calculation, the initial value 0, j is a subscript and is max Zij；
Because the convolution operation is linear operation, the nonlinear processing is carried out by using a loss function, the loss function uses pixel-wise softmax, the output corresponding to the pixel is independently made into softmax, and the formula is as follows:
in the formula, x is a pixel position on a two-dimensional plane, a is a learning coefficient, an initial value is 1, w (x) is a weight term in cross entropy, pl (x) represents the output probability of x on a channel where a real label is located, c is a constant term, and the initial value is 0;
5-2) the downsampling operation is as follows: sending the image with 6045 pixels obtained by convolution operation into a Decoder network, increasing the length and width of the image into 12090 pixels by two times after upsampling operation, repeating the two times of the sampling operation on the Decoder network to obtain 480360 pixels, performing convolution operation by using 55 convolution kernels to obtain 420300 pixels of the image, performing convolution operation by using 51 convolution kernels to obtain 400300 pixels of the image, and performing upsampling operation once to recover 800600 pixels which are the same as the original image;
5-3) carrying out full connection operation on the recovered 800600 acorn image to obtain the specific position of a single character in the original image, and expressing the position by using the upper left corner coordinate and the lower right corner coordinate of the sub-image;
the full join operation is as follows: the positions of 8 character images are mainly output by full connection, each image coordinate consists of 4 values in the upper left corner and the lower right corner, and the 8 images have 32 output values; the image is a two-dimensional array of 800600, which is converted into a one-dimensional array by rows with length of 480000, then 32 one-dimensional array parameters with length of 48000 are multiplied and summed with the image one-dimensional array pixel value, and an intercept parameter is added, so that the obtained 32 values are the coordinate positions of 8 characters;
full connection formula: x is the number ofi＝anm*wi+ci
In the formula xiIs 32 coordinate values, i takes a value from 1 to 32, anmIs a one-dimensional array of images, wiIs a one-dimensional array parameter, length n m, ciIs an intercept parameter, wiAnd ciAre both learnable parameters, n and m are the length and width of the source image, respectively, i.e. n is 800 and m is 600;
6) classification and identification: respectively subjecting the sub-images which are segmented in the step 5) and contain single characters to convolutional neural network to sequential classification and recognition, wherein each sub-image is recognized once to obtain a character value and a credible value, and 8 classification and recognition are carried out for 8 times to obtain 8 character values and 8 corresponding credible values, and each credible value is larger than 90%;
the convolutional neural network structure has 8 layers, wherein: the 1-3 layer network uses 9 types of convolution kernels to extract 9 characteristics, each type of convolution kernel is 33, the 4-6 layer network uses 12 types of convolution kernels, each convolution kernel is 33, the 7 th layer of convolution kernel uses 1024 types of convolution kernels, each convolution kernel is 11, the 8 th layer is a full-connection layer, 62 credible values are output, and the character corresponding to the position with the maximum credible value is the recognized character value;
after each convolution operation, carrying out nonlinear operation by using an activation function, wherein the activation function uses an exponential linear unit E L U function;
7) the obtained 8 character values are transmitted to a production scheduling center in real time through a wireless network, the character numbers are used as query conditions, the specific positions are looked up in a database system, the underground positions of the unmanned locomotives can be determined, the underground positioning of the unmanned locomotives in the mine is achieved, meanwhile, image data with low probability values of recognition results are automatically stored, the stored images are trained in a targeted mode every month, the training rate is 99.89%, the network model on production is updated, and the purpose of continuous learning is achieved.
Claims (1)
1. A method for obtaining the position of underground equipment of a mine based on character image intelligent identification is characterized by comprising the following steps:
1) installing a plurality of character boards beside a mine underground rail at intervals, marking a plurality of characters on each character board, and recording the rail position corresponding to each character number in a database of a production scheduling center;
2) installing image acquisition equipment on an unmanned locomotive under a mine, and acquiring all character images marked on corresponding character boards during running;
3) reading the acquired character image data by using a video Capture class in an open source image processing library OpenCV, wherein the frame rate is 15 frames per second, the pixel format is an RGB three-channel image, and the original pixel is 1024900;
4) carrying out conventional zooming and filtering pretreatment on each frame of image data, wherein the zooming treatment is to zoom or enlarge the image size so as to reduce the processing amount of deep learning network model data and accelerate the segmentation and identification speed of each frame of image, and the L anczos algorithm is used for compressing the image into 800600 pixels in the form of RGB color image;
5) the preprocessed 800600-pixel image is delivered to a U-Net network for detection, after 8-character images in the image are detected, the image is divided into 8 non-overlapping sub-images, and each sub-image contains 1 character;
the U-Net network is an Encoder-Decoder network structure, wherein: the method comprises the following steps that an Encoder network is used for convolution operation, a Decode network is used for up-sampling operation, the Encoder network is of a five-layer convolution network structure, the convolution kernel of each layer of convolution network structure is 55, padding is 0, striping is 1, the Decode network is of a five-layer convolution network structure, the convolution kernel of each layer of convolution network is 11, and the step length is 1;
the convolution operation and the downsampling operation are performed using the following algorithm:
5-1) convolution operation as follows: processing 800600 pixel image into 780580 pixel by five-layer convolution operation, and pooling with step size 2 to 390290 pixel by 22 convolution kernel; repeating the operation for three times to obtain a pixel with an image of 6045; the convolution operation is formulated as follows:
where X is the image data, i and j are the image sizes, 800 and 600 respectively, W is the convolution kernel, m and n are the convolution kernel sizes, here 5 and 5 respectively, and s (i, j) is the new image data after the convolution operation;
after each convolution operation, performing nonlinear calculation by using an activation function, wherein the activation function of the whole network uses a Maxout activation function, and the activation function formula is as follows:
wherein
In the formula xTIs the network neuron number, WijIs convolution kernel value, i and j are coordinate positions in the convolution kernel, k is the number of channels of the image, the image is an RGB color image, k is 3, bij is a constant corresponding to each neuron, c is an empirical constant after activation calculation, the initial value 0, j is a subscript and is max Zij；
Because the convolution operation is linear operation, the nonlinear processing is carried out by using a loss function, the loss function uses pixel-wise softmax, the output corresponding to the pixel is independently made into softmax, and the formula is as follows:
in the formula, x is a pixel position on a two-dimensional plane, a is a learning coefficient, an initial value is 1, w (x) is a weight term in cross entropy, pl (x) represents the output probability of x on a channel where a real label is located, c is a constant term, and the initial value is 0;
5-2) the downsampling operation is as follows: sending the image with 6045 pixels obtained by convolution operation into a Decoder network, increasing the length and width of the image into 12090 pixels by two times after upsampling operation, repeating the two times of the sampling operation on the Decoder network to obtain 480360 pixels, performing convolution operation by using 55 convolution kernels to obtain 420300 pixels of the image, performing convolution operation by using 51 convolution kernels to obtain 400300 pixels of the image, and performing upsampling operation once to recover 800600 pixels which are the same as the original image;
5-3) carrying out full connection operation on the recovered 800600 acorn image to obtain the specific position of a single character in the original image, and expressing the position by using the upper left corner coordinate and the lower right corner coordinate of the sub-image;
the full join operation is as follows: the positions of 8 character images are mainly output by full connection, each image coordinate consists of 4 values in the upper left corner and the lower right corner, and the 8 images have 32 output values; the image is a two-dimensional array of 800600, which is converted into a one-dimensional array by rows with length of 480000, then 32 one-dimensional array parameters with length of 48000 are multiplied and summed with the image one-dimensional array pixel value, and an intercept parameter is added, so that the obtained 32 values are the coordinate positions of 8 characters;
full connection formula:
in the formula xiIs 32 coordinate values, i takes a value from 1 to 32, anmIs a one-dimensional array of images, wiIs a one-dimensional array parameter, length n m, ciIs an intercept parameter, wiAnd ciAre both learnable parameters, n and m are the length and width of the source image, respectively, i.e. n is 800 and m is 600;
6) classification and identification: respectively subjecting the sub-images which are segmented in the step 5) and contain single characters to convolutional neural network to sequential classification and recognition, wherein each sub-image is recognized once to obtain a character value and a credible value, and 8 classification and recognition are carried out for 8 times to obtain 8 character values and 8 corresponding credible values, and each credible value is larger than 90%;
the convolutional neural network structure has 8 layers, wherein: the 1-3 layer network uses 9 types of convolution kernels to extract 9 characteristics, each type of convolution kernel is 33, the 4-6 layer network uses 12 types of convolution kernels, each convolution kernel is 33, the 7 th layer of convolution kernel uses 1024 types of convolution kernels, each convolution kernel is 11, the 8 th layer is a full-connection layer, 62 credible values are output, and the character corresponding to the position with the maximum credible value is the recognized character value;
after each convolution operation, carrying out nonlinear operation by using an activation function, wherein the activation function uses an exponential linear unit E L U function;
7) the obtained 8 character values are transmitted to a production scheduling center in real time through a wireless network, the character numbers are used as query conditions, the specific positions are looked up in a database system, the underground positions of the unmanned locomotives can be determined, the underground positioning of the unmanned locomotives in the mine is achieved, meanwhile, image data with low probability values of recognition results are automatically stored, the stored images are trained in a targeted mode every month, the training rate is 99.89%, the network model on production is updated, and the purpose of continuous learning is achieved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010114364.8A CN111401139B (en) | 2020-02-25 | 2020-02-25 | Method for obtaining mine underground equipment position based on character image intelligent recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010114364.8A CN111401139B (en) | 2020-02-25 | 2020-02-25 | Method for obtaining mine underground equipment position based on character image intelligent recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111401139A true CN111401139A (en) | 2020-07-10 |
| CN111401139B CN111401139B (en) | 2024-03-29 |
Family
ID=71430423
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010114364.8A Active CN111401139B (en) | 2020-02-25 | 2020-02-25 | Method for obtaining mine underground equipment position based on character image intelligent recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111401139B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112738767A (en) * | 2020-11-30 | 2021-04-30 | 中南大学 | Trust-based mobile edge user task scheduling method |
| CN113112431A (en) * | 2021-05-10 | 2021-07-13 | 苏州大学 | Image processing method in embedded system |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446895A (en) * | 2016-10-28 | 2017-02-22 | 安徽四创电子股份有限公司 | License plate recognition method based on deep convolutional neural network |
| CN106650721A (en) * | 2016-12-28 | 2017-05-10 | 吴晓军 | Industrial character identification method based on convolution neural network |
| CN107688784A (en) * | 2017-08-23 | 2018-02-13 | 福建六壬网安股份有限公司 | A kind of character identifying method and storage medium based on further feature and shallow-layer Fusion Features |
| CN107967475A (en) * | 2017-11-16 | 2018-04-27 | 广州探迹科技有限公司 | A kind of method for recognizing verification code based on window sliding and convolutional neural networks |
| CN108108746A (en) * | 2017-09-13 | 2018-06-01 | 湖南理工学院 | License plate character recognition method based on Caffe deep learning frames |
| CN109344825A (en) * | 2018-09-14 | 2019-02-15 | 广州麦仑信息科技有限公司 | A kind of licence plate recognition method based on convolutional neural networks |
| CN109740603A (en) * | 2019-01-21 | 2019-05-10 | 闽江学院 | Based on the vehicle character identifying method under CNN convolutional neural networks |
| CN110414506A (en) * | 2019-07-04 | 2019-11-05 | 南京理工大学 | Bank card number automatic identifying method based on data augmentation and convolutional neural networks |
| CN110619329A (en) * | 2019-09-03 | 2019-12-27 | 中国矿业大学 | Carriage number and loading state identification method of railway freight open wagon based on airborne vision |
| CN110766002A (en) * | 2019-10-08 | 2020-02-07 | 浙江大学 | Ship name character region detection method based on deep learning |
-
2020
- 2020-02-25 CN CN202010114364.8A patent/CN111401139B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446895A (en) * | 2016-10-28 | 2017-02-22 | 安徽四创电子股份有限公司 | License plate recognition method based on deep convolutional neural network |
| CN106650721A (en) * | 2016-12-28 | 2017-05-10 | 吴晓军 | Industrial character identification method based on convolution neural network |
| CN107688784A (en) * | 2017-08-23 | 2018-02-13 | 福建六壬网安股份有限公司 | A kind of character identifying method and storage medium based on further feature and shallow-layer Fusion Features |
| CN108108746A (en) * | 2017-09-13 | 2018-06-01 | 湖南理工学院 | License plate character recognition method based on Caffe deep learning frames |
| CN107967475A (en) * | 2017-11-16 | 2018-04-27 | 广州探迹科技有限公司 | A kind of method for recognizing verification code based on window sliding and convolutional neural networks |
| CN109344825A (en) * | 2018-09-14 | 2019-02-15 | 广州麦仑信息科技有限公司 | A kind of licence plate recognition method based on convolutional neural networks |
| CN109740603A (en) * | 2019-01-21 | 2019-05-10 | 闽江学院 | Based on the vehicle character identifying method under CNN convolutional neural networks |
| CN110414506A (en) * | 2019-07-04 | 2019-11-05 | 南京理工大学 | Bank card number automatic identifying method based on data augmentation and convolutional neural networks |
| CN110619329A (en) * | 2019-09-03 | 2019-12-27 | 中国矿业大学 | Carriage number and loading state identification method of railway freight open wagon based on airborne vision |
| CN110766002A (en) * | 2019-10-08 | 2020-02-07 | 浙江大学 | Ship name character region detection method based on deep learning |
Non-Patent Citations (2)
| Title |
|---|
| 欧先锋 等: "《基于CNN的车牌数字字符识别算法》", 《成都工业学院学报》, vol. 19, no. 4, 31 December 2016 (2016-12-31) * |
| 赵成龙: "《复杂自然环境下车牌识别算法》", 《中国优秀硕士学位论文全文数据库 信息科技辑》, 15 January 2018 (2018-01-15) * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112738767A (en) * | 2020-11-30 | 2021-04-30 | 中南大学 | Trust-based mobile edge user task scheduling method |
| CN113112431A (en) * | 2021-05-10 | 2021-07-13 | 苏州大学 | Image processing method in embedded system |
| WO2022237062A1 (en) * | 2021-05-10 | 2022-11-17 | 苏州大学 | Image processing method in embedded system |
| US11622169B1 (en) | 2021-05-10 | 2023-04-04 | Soochow University | Picture processing method in embedded system |
| CN113112431B (en) * | 2021-05-10 | 2023-08-15 | 苏州大学 | Image processing method in embedded system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111401139B (en) | 2024-03-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111080620B (en) | Road disease detection method based on deep learning | |
| Wang et al. | Deep learning for asphalt pavement cracking recognition using convolutional neural network | |
| CN109255284B (en) | Motion trajectory-based behavior identification method of 3D convolutional neural network | |
| CN112381788B (en) | Part surface defect increment detection method based on double-branch matching network | |
| CN110648310B (en) | Weak supervision casting defect identification method based on attention mechanism | |
| CN107316016A (en) | A kind of track of vehicle statistical method based on Hadoop and monitoring video flow | |
| CN104992223A (en) | Dense population estimation method based on deep learning | |
| CN113324864B (en) | Pantograph carbon slide plate abrasion detection method based on deep learning target detection | |
| CN111401139B (en) | Method for obtaining mine underground equipment position based on character image intelligent recognition | |
| CN111008632B (en) | License plate character segmentation method based on deep learning | |
| CN108198417B (en) | A kind of road cruising inspection system based on unmanned plane | |
| CN115311241B (en) | Underground coal mine pedestrian detection method based on image fusion and feature enhancement | |
| CN112966665A (en) | Pavement disease detection model training method and device and computer equipment | |
| CN113160575A (en) | Traffic violation detection method and system for non-motor vehicles and drivers | |
| CN112967252B (en) | Rail vehicle machine sense hanger assembly bolt loss detection method | |
| CN111598855A (en) | 2C equipment high-speed rail contact net dropper defect detection method based on deep learning and transfer learning | |
| Shanthakumari et al. | Mask RCNN and Tesseract OCR for vehicle plate character recognition | |
| CN111105396A (en) | Printed matter quality detection method and system based on artificial intelligence | |
| CN111881914B (en) | License plate character segmentation method and system based on self-learning threshold | |
| CN112053407B (en) | Automatic lane line detection method based on AI technology in traffic law enforcement image | |
| CN115147450B (en) | Moving target detection method and detection device based on motion frame difference image | |
| CN108734158B (en) | Real-time train number identification method and device | |
| CN112115767B (en) | Tunnel foreign matter detection method based on Retinex and YOLOv3 models | |
| CN113192018B (en) | Water-cooled wall surface defect video identification method based on fast segmentation convolutional neural network | |
| CN115565141A (en) | Truck axle type detection method based on visual infrared fusion |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |