CN113963193A - Method and device for generating vehicle body color classification model and storage medium - Google Patents
Method and device for generating vehicle body color classification model and storage medium Download PDFInfo
- Publication number
- CN113963193A CN113963193A CN202111109539.7A CN202111109539A CN113963193A CN 113963193 A CN113963193 A CN 113963193A CN 202111109539 A CN202111109539 A CN 202111109539A CN 113963193 A CN113963193 A CN 113963193A
- Authority
- CN
- China
- Prior art keywords
- vehicle body
- color classification
- classification model
- image
- body color
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- 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
Abstract
The invention discloses a method, a device and a storage medium for generating a vehicle body color classification model, wherein the method comprises the following steps: acquiring a vehicle body image sample set; the vehicle body image sample set comprises vehicle body images shot from multiple angles in multiple scenes; acquiring a label for labeling each vehicle body image according to a preset color classification standard; the preset color classification standard is used for indicating the actual color of the vehicle body in various environments; converting image data corresponding to each vehicle body image into frequency domain data; the image data corresponding to the vehicle body image is time domain data; dividing frequency domain data corresponding to the vehicle body image sample set into a training set and a test set according to a certain proportion of each vehicle body color category; and training the convolutional neural network model to be trained by adopting the training sample set and the labels corresponding to the training sample set until the classification precision and the loss value tend to be stable, and obtaining the vehicle body color classification model. The invention can improve the effect of vehicle body color classification.
Description
Technical Field
The invention relates to the technical field of computer vision, in particular to a method and a device for generating a vehicle body color classification model and a storage medium.
Background
Vehicle body identification is an important aid in vehicle identification systems. Body color recognition also plays an important role in traffic research and traffic management. At present, the vehicle body identification method mainly comprises the following three prior arts:
the prior art 1 mainly extracts an effective region of interest, obtains a value of an image mapping Lab color space combined with an RGB space through affine transformation, trains by using a neural network model, and classifies vehicle colors by combining a support vector machine. However, the inventor finds that, since the method of the prior art 1 does not consider that the color in the Lab space is the output modeled on the physical device, and not the visual perception of human, the color labeling thereof after RGB conversion depends on the device and human subjectivity, and there is no absolute reference value relative to the standard, and the result of processing the same color by different devices may be calibrated to different colors.
In the prior art 2, different regions of the whole vehicle are marked into a main clue region, an auxiliary recognition region and a non-recognition region, then a CNN model is used for training, and different regions are detected and assigned with different weights for color classification. However, the inventor finds that the method of the prior art 2 does not consider that many vehicle body images cannot obtain the main line cable region, the auxiliary recognition region and the unrecognizable region due to large-angle shooting or shooting under different scene conditions, and a certain region of the same vehicle may be an unrecognizable region or a main line cable region under different conditions.
Prior art 3 mainly includes obtaining all regions to be identified of a vehicle, including at least one sub-ROI region, classifying by using a preset classifier, matching the features extracted by the ROI with the preset features, and selecting the highest confidence as the classification result. However, the inventor finds that in the prior art 3, extracted ROI features are fused and matched with a preset classifier, the confidence of selection is high, but interference caused by dust and the like is ignored, color transformation is in a nonlinear relation, and misjudgment is large during classification.
Disclosure of Invention
The embodiment of the invention provides a method, a device and a storage medium for generating a vehicle body color classification model, which are used for solving the technical problems that in the prior art, after the same color is converted, the same color is marked into multiple colors, the vehicle body color classification model depends on a specific vehicle body subregion, and the vehicle body color classification effect is poor due to the fact that the interference environment is not considered.
In a first aspect, an embodiment of the present invention provides a method for generating a vehicle body color classification model, including:
acquiring a vehicle body image sample set; the vehicle body image sample set comprises vehicle body images shot from multiple angles in multiple scenes;
acquiring a label for labeling each vehicle body image according to a preset color classification standard; the preset color classification standard is used for indicating the actual color of the vehicle body in various environments;
converting image data corresponding to each vehicle body image into frequency domain data; the image data corresponding to the vehicle body image is time domain data;
dividing frequency domain data corresponding to the vehicle body image sample set into a training set and a test set according to a certain proportion of each vehicle body color category;
and training the convolutional neural network model to be trained by adopting the training sample set and the labels corresponding to the training sample set until the classification precision and the loss value tend to be stable, and obtaining the vehicle body color classification model.
Preferably, after obtaining the vehicle body color classification model, the method further includes:
acquiring a standard test set; the standard test set is obtained by performing time-frequency conversion on effective samples;
testing the vehicle body color classification model by adopting the standard test set to obtain a test result;
calculating the classification precision of the vehicle body color classification model according to the test result and the actual result;
and if the classification precision does not meet the preset requirement, acquiring the test set, correcting part of weights in the vehicle body color classification model by using the test set, and regenerating the vehicle body color classification model.
Preferably, after obtaining the regenerated vehicle body color classification model, the method further comprises:
testing the regenerated vehicle body color classification model by adopting the standard test set;
and if the classification precision of the regenerated vehicle body color classification model does not meet the preset requirement, outputting information for requesting to modify the model architecture.
Preferably, the method for converting image data corresponding to a vehicle body image into frequency domain data comprises:
transforming the vehicle body image into a multi-dimensional matrix of 3 × H × W; wherein H, W represents the height and width of the vehicle body image, respectively;
converting the multidimensional matrix into a frequency domain space by adopting discrete Fourier transform to obtain a discrete complex matrix; and the discrete complex matrix is frequency domain data corresponding to the vehicle body image.
Preferably, the method for converting the image data corresponding to the image of the vehicle body into the frequency domain data comprises
And converting the image data corresponding to the vehicle body image into frequency domain data by adopting wavelet transformation.
Preferably, the plurality of scenes includes, but is not limited to, snow, dust and clean scenes.
In a second aspect, an embodiment of the present invention provides a vehicle body color classification model generation apparatus, including:
the first acquisition unit is used for acquiring a vehicle body image sample set; the vehicle body image sample set comprises vehicle body images shot from multiple angles in multiple scenes;
the second acquisition unit is used for acquiring a label for labeling each vehicle body image according to a preset color classification standard; the preset color classification standard is used for indicating the actual color of the vehicle body in various environments;
the time-frequency conversion unit is used for converting the image data corresponding to each vehicle body image into frequency domain data; the image data corresponding to the vehicle body image is time domain data;
the dividing unit is used for dividing the frequency domain data corresponding to the vehicle body image sample set into a training set and a test set according to a certain proportion of each vehicle body color category;
and the training unit is used for training the convolutional neural network model to be trained by adopting the training sample set and the labels corresponding to the training sample set until the classification precision and the loss value tend to be stable, and obtaining the vehicle body color classification model.
Preferably, the vehicle body color classification model generation device includes:
a third obtaining unit, configured to obtain a standard test set; the standard test set is obtained by performing time-frequency conversion on effective samples;
the test unit is used for testing the vehicle body color classification model by adopting the standard test set to obtain a test result;
the classification precision calculation unit is used for calculating the classification precision of the vehicle body color classification model according to the test result and the actual result;
and the correcting unit is used for acquiring the test set if the classification precision does not meet the preset requirement, correcting part of weights in the vehicle body color classification model by using the test set, and regenerating the vehicle body color classification model.
In a third aspect, an embodiment of the present invention further provides an electronic device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, the processor implements the method for generating the vehicle body color classification model according to any one of the above items.
In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium includes a stored computer program, where when the computer program runs, the apparatus in which the computer-readable storage medium is located is controlled to perform the method for generating the vehicle body color classification model according to any one of the above items.
Compared with the prior art, the method for generating the vehicle body color classification model disclosed by the embodiment of the invention comprises the following steps: acquiring a vehicle body image sample set; the vehicle body image sample set comprises vehicle body images shot from multiple angles in multiple scenes; acquiring a label for labeling each vehicle body image according to a preset color classification standard; the preset color classification standard is used for indicating the actual color of the vehicle body in various environments; converting image data corresponding to each vehicle body image into frequency domain data; the image data corresponding to the vehicle body image is time domain data; dividing frequency domain data corresponding to the vehicle body image sample set into a training set and a test set according to a certain proportion of each vehicle body color category; and training the convolutional neural network model to be trained by adopting the training sample set and the labels corresponding to the training sample set until the classification precision and the loss value tend to be stable, and obtaining the vehicle body color classification model. According to the embodiment of the invention, a color classification standard for indicating the actual colors of the vehicle body in various environments is drawn up on the time domain RGB color space, so that the condition that the same color is labeled as multiple colors after being transformed does not occur, and all vehicle images can realize actual color classification, thereby improving the vehicle body color classification effect. In addition, the embodiment of the invention converts the original image from the time domain to the frequency domain, and can separate the interference signal and the color characteristic signal caused by the environmental factors, thereby reducing the influence of the interference environment on the color classification effect of the vehicle body and further improving the color classification effect of the vehicle body. In addition, the embodiment of the invention adopts the convolutional neural network model, and can realize the elimination of absolute invalid regions such as vehicle windows and license plates, thereby reducing the influence of the invalid regions on the color classification effect of the vehicle body and further improving the color classification effect of the vehicle body.
Drawings
FIG. 1 is a detailed flow chart of a method for generating a color classification model of a vehicle body according to an embodiment of the invention;
FIG. 2 is a simplified flow diagram of a method for generating a color classification model of a vehicle body according to an embodiment of the present invention;
FIG. 3 is a training process for a convolutional neural network;
fig. 4 is a schematic structural diagram of a vehicle body color classification model generation device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1 and 2, a method for generating a vehicle body color classification model according to an embodiment of the present invention includes steps S1-S5:
s1, obtaining a vehicle body image sample set; the body image sample set comprises body images shot from multiple angles in multiple scenes.
In the embodiment of the present invention, the vehicle body images in the vehicle body image sample set are vehicle body images taken at a plurality of angles, for example, only see a vehicle face, look down the entire vehicle body, and the like, in a plurality of scenes, for example, snow, dust, clean scenes, and the like.
S2, acquiring labels for labeling each vehicle body image according to preset color classification standards; the preset color classification standard is used for indicating the actual color of the vehicle body in various environments.
In the embodiment of the present invention, it should be noted that the actual color of the vehicle body does not change with the change and variation of the environment, however, the environment may cause the person to distinguish the color of the vehicle body. Since the actual color of the vehicle plays an important role in traffic investigation and traffic management, in order to identify the actual body color of the vehicle in various environments, the embodiment of the present invention sets a classification standard for indicating the actual color of the body in various environments. Specifically, the preset color classification standard includes a unified standard under interference conditions such as no interference, snow dust light, and the like.
S3, converting the image data corresponding to each vehicle body image into frequency domain data; and the image data corresponding to the vehicle body image is time domain data.
It should be noted that the frequency domain does not exist really, but a mapping space constructed by mathematical theory is usually a coordinate space used for describing the characteristics of the signal in terms of frequency, and the time domain is the only domain existing objectively. Many signals are similar in the time domain, but after mapping to a frequency domain space, the difference is large and can be easily distinguished. The main application of time-domain conversion to frequency domain is fourier transform, i.e. any discrete signal is converted into a signal in the field of mathematical construction by discrete fourier transform, and expressed with the frequency axis as a coordinate.
In addition to using discrete fourier transform to convert image time domain data into frequency domain data, a more versatile approach is to apply wavelet transform. The wavelet transform can divide the signal data into high frequency and low frequency, so as to facilitate various filtering processes, for example, for the color information of the vehicle body, the main color of the vehicle body belongs to the low frequency information, and other interference noise information belongs to the high frequency vibration information, and the way of filtering the high frequency interference information can replace the smoothing filtering way in the traditional image processing.
And S4, dividing the frequency domain data corresponding to the vehicle body image sample set into a training set and a test set according to a certain proportion of each vehicle body color type.
And S5, training the convolutional neural network model to be trained by adopting the training sample set and the labels corresponding to the training sample set until the classification precision and the loss value tend to be stable, and obtaining the vehicle body color classification model.
In the embodiment of the present invention, it should be understood that, the label corresponding to the training sample set is a label corresponding to each vehicle image in the training sample set, and the label identifies a vehicle body color of the vehicle image.
In an embodiment of the present invention, the training process of the Convolutional Neural Network (CNN) is a one-way network. The method mainly comprises the steps of amplifying and amplifying characteristic data of a frequency domain, then extracting an ROI (region of interest) through operations such as general convolution, pooling and the like, and taking other regions after invalid regions are removed as the ROI. The features of the ROI region are fused and then classified using a classifier. And if the classification result is not matched with the label, reversely calculating the descending gradient, repeating the iterative training until the fitting capacity of the model is stronger, realizing the correct classification of the colors of the vehicle body, and then outputting the result as a test model. The specific flow is shown in fig. 3.
According to the embodiment of the invention, a color classification standard for indicating the actual colors of the vehicle body in various environments is drawn up on the time domain RGB color space, so that the condition that the same color is labeled as multiple colors after being transformed does not occur, and all vehicle images can realize actual color classification, thereby improving the vehicle body color classification effect. In addition, the embodiment of the invention converts the original image from the time domain to the frequency domain, and can separate the interference signal and the color characteristic signal caused by the environmental factors, thereby reducing the influence of the interference environment on the color classification effect of the vehicle body and further improving the color classification effect of the vehicle body. In addition, the embodiment of the invention adopts the convolutional neural network model, and can realize the elimination of absolute invalid regions such as vehicle windows and license plates, thereby reducing the influence of the invalid regions on the color classification effect of the vehicle body and further improving the color classification effect of the vehicle body.
As an example of the embodiment of the present invention, after obtaining the vehicle body color classification model, the method further includes:
s6, acquiring a standard test set; and the standard test set is obtained by performing time-frequency conversion on the effective samples.
In the embodiment of the invention, when needing to be explained, each sample in the standard test set is correctly classified manually, so that the classification accuracy of the vehicle body color classification model is verified conveniently.
S7, testing the vehicle body color classification model by adopting the standard test set to obtain a test result;
s8, calculating the classification precision of the vehicle body color classification model according to the test result and the actual result;
and S9, if the classification precision does not meet the preset requirement, acquiring the test set, correcting part of weights in the vehicle body color classification model by using the test set, and regenerating the vehicle body color classification model.
According to the embodiment of the invention, when the classification precision of the initially generated vehicle body color classification model does not meet the preset requirement, the test set is adopted to correct part of the weight of the initially generated vehicle body color classification model, so that the vehicle body color classification model with higher classification precision can be obtained.
As an example of the embodiment of the present invention, after obtaining the regenerated vehicle body color classification model, the method further includes:
testing the regenerated vehicle body color classification model by adopting the standard test set;
and if the classification precision of the regenerated vehicle body color classification model does not meet the preset requirement, outputting information for requesting to modify the model architecture.
According to the embodiment of the invention, when the classification accuracy of the regenerated vehicle body color classification model still does not meet the preset requirement, the information of modifying the model architecture is output, so that corresponding research personnel can be reminded to modify the model architecture in time.
As an example of the embodiment of the present invention, a method for converting image data corresponding to a vehicle body image into frequency domain data includes:
transforming the vehicle body image into a multi-dimensional matrix of 3 × H × W; wherein H, W represents the height and width of the vehicle body image, respectively;
converting the multidimensional matrix into a frequency domain space by adopting discrete Fourier transform to obtain a discrete complex matrix; and the discrete complex matrix is frequency domain data corresponding to the vehicle body image.
As an example of the embodiment of the invention, the method for converting the image data corresponding to the vehicle body image into the frequency domain data comprises the following steps
And converting the image data corresponding to the vehicle body image into frequency domain data by adopting wavelet transformation.
Example 2:
referring to fig. 4, an embodiment of the present invention provides a device for generating a vehicle body color classification model, including:
a first acquisition unit 1 for acquiring a body image sample set; the vehicle body image sample set comprises vehicle body images shot from multiple angles in multiple scenes;
the second obtaining unit 2 is used for labeling each car body image according to a preset color classification standard; the preset color classification standard is used for indicating the actual color of the vehicle body in various environments; (ii) a
The time-frequency conversion unit 3 is used for converting the image data corresponding to each vehicle body image into frequency domain data; the image data corresponding to the vehicle body image is time domain data;
the dividing unit 4 is used for dividing the frequency domain data corresponding to the vehicle body image sample set into a training set and a test set according to a certain proportion of each vehicle body color class;
and the training unit 5 is used for training the convolutional neural network model to be trained by adopting the training sample set and the corresponding labels until the classification precision and the loss value tend to be stable, and then stopping training to obtain the vehicle body color classification model.
As an example of the embodiment of the present invention, the vehicle body color classification model generation apparatus further includes:
a third obtaining unit, configured to obtain a standard test set; the standard test set is obtained by performing time-frequency conversion on effective samples;
the test unit is used for testing the vehicle body color classification model by adopting the standard test set to obtain a test result;
the classification precision calculation unit is used for calculating the classification precision of the vehicle body color classification model according to the test result and the actual result;
and the correcting unit is used for acquiring the test set if the classification precision does not meet the preset requirement, correcting part of weights in the vehicle body color classification model by using the test set, and regenerating the vehicle body color classification model.
Example 3:
an embodiment of the present invention provides an electronic device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, the processor implements the method for generating the vehicle body color classification model according to any one of the above embodiments.
Example 4:
the embodiment of the invention provides a computer-readable storage medium, which comprises a stored computer program, wherein when the computer program runs, a device where the computer-readable storage medium is located is controlled to execute the method for generating the vehicle body color classification model according to any one of the above items.
It should be noted that, all or part of the flow in the method according to the above embodiments of the present invention may also be implemented by a computer program instructing related hardware, where the computer program may be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above embodiments of the method may be implemented. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like. It should be further noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals in accordance with legislation and patent practice.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.
Claims (10)
1. A method of vehicle body color classification model generation, comprising:
acquiring a vehicle body image sample set; the vehicle body image sample set comprises vehicle body images shot from multiple angles in multiple scenes;
acquiring a label for labeling each vehicle body image according to a preset color classification standard; the preset color classification standard is used for indicating the actual color of the vehicle body in various environments;
converting image data corresponding to each vehicle body image into frequency domain data; the image data corresponding to the vehicle body image is time domain data;
dividing frequency domain data corresponding to the vehicle body image sample set into a training set and a test set according to a certain proportion of each vehicle body color category;
and training the convolutional neural network model to be trained by adopting the training sample set and the labels corresponding to the training sample set until the classification precision and the loss value tend to be stable, and obtaining the vehicle body color classification model.
2. The method for generating the vehicle body color classification model according to claim 1, further comprising, after obtaining the vehicle body color classification model:
acquiring a standard test set; the standard test set is obtained by performing time-frequency conversion on effective samples;
testing the vehicle body color classification model by adopting the standard test set to obtain a test result;
calculating the classification precision of the vehicle body color classification model according to the test result and the actual result;
and if the classification precision does not meet the preset requirement, acquiring the test set, correcting part of weights in the vehicle body color classification model by using the test set, and regenerating the vehicle body color classification model.
3. The method for generating a body color classification model according to claim 2, further comprising, after obtaining the regenerated body color classification model:
testing the regenerated vehicle body color classification model by adopting the standard test set;
and if the classification precision of the regenerated vehicle body color classification model does not meet the preset requirement, outputting information for requesting to modify the model architecture.
4. The method for generating the vehicle body color classification model according to any one of claims 1 to 3, wherein the method for converting the image data corresponding to one vehicle body image into the frequency domain data comprises the following steps:
transforming the vehicle body image into a multi-dimensional matrix of 3 × H × W; wherein H, W represents the height and width of the vehicle body image, respectively;
converting the multidimensional matrix into a frequency domain space by adopting discrete Fourier transform to obtain a discrete complex matrix; and the discrete complex matrix is frequency domain data corresponding to the vehicle body image.
5. The method for generating the vehicle body color classification model according to any one of claims 1 to 3, wherein the method for converting the image data corresponding to one vehicle body image into the frequency domain data comprises the following steps:
and converting the image data corresponding to the vehicle body image into frequency domain data by adopting wavelet transformation.
6. The method for body color classification model generation according to claim 1, characterized in that the scenes include, but are not limited to, snow, dust and clean scenes.
7. A vehicle body color classification model generation device is characterized by comprising:
the first acquisition unit is used for acquiring a vehicle body image sample set; the vehicle body image sample set comprises vehicle body images shot from multiple angles in multiple scenes;
the second acquisition unit is used for acquiring a label for labeling each vehicle body image according to a preset color classification standard; the preset color classification standard is used for indicating the actual color of the vehicle body in various environments;
the time-frequency conversion unit is used for converting the image data corresponding to each vehicle body image into frequency domain data; the image data corresponding to the vehicle body image is time domain data;
the dividing unit is used for dividing the frequency domain data corresponding to the vehicle body image sample set into a training set and a test set according to a certain proportion of each vehicle body color category;
and the training unit is used for training the convolutional neural network model to be trained by adopting the training sample set and the labels corresponding to the training sample set until the classification precision and the loss value tend to be stable, and obtaining the vehicle body color classification model.
8. The vehicle body color classification model generation apparatus according to claim 7, characterized by further comprising:
a third obtaining unit, configured to obtain a standard test set; the standard test set is obtained by performing time-frequency conversion on effective samples;
the test unit is used for testing the vehicle body color classification model by adopting the standard test set to obtain a test result;
the classification precision calculation unit is used for calculating the classification precision of the vehicle body color classification model according to the test result and the actual result;
and the correcting unit is used for acquiring the test set if the classification precision does not meet the preset requirement, correcting part of weights in the vehicle body color classification model by using the test set, and regenerating the vehicle body color classification model.
9. An electronic device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the method of vehicle body color classification model generation as claimed in any one of claims 1 to 6 when executing the computer program.
10. A computer-readable storage medium, comprising a stored computer program, wherein the computer program, when executed, controls an apparatus in which the computer-readable storage medium is located to perform the method of generating a body color classification model according to any one of claims 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111109539.7A CN113963193A (en) | 2021-09-22 | 2021-09-22 | Method and device for generating vehicle body color classification model and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111109539.7A CN113963193A (en) | 2021-09-22 | 2021-09-22 | Method and device for generating vehicle body color classification model and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113963193A true CN113963193A (en) | 2022-01-21 |
Family
ID=79462163
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111109539.7A Pending CN113963193A (en) | 2021-09-22 | 2021-09-22 | Method and device for generating vehicle body color classification model and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113963193A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115410048A (en) * | 2022-09-29 | 2022-11-29 | 昆仑芯(北京)科技有限公司 | Training method, device, equipment and medium of image classification model and image classification method, device and equipment |
CN116665138A (en) * | 2023-08-01 | 2023-08-29 | 临朐弘泰汽车配件有限公司 | Visual detection method and system for stamping processing of automobile parts |
-
2021
- 2021-09-22 CN CN202111109539.7A patent/CN113963193A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115410048A (en) * | 2022-09-29 | 2022-11-29 | 昆仑芯(北京)科技有限公司 | Training method, device, equipment and medium of image classification model and image classification method, device and equipment |
CN115410048B (en) * | 2022-09-29 | 2024-03-19 | 昆仑芯(北京)科技有限公司 | Training of image classification model, image classification method, device, equipment and medium |
CN116665138A (en) * | 2023-08-01 | 2023-08-29 | 临朐弘泰汽车配件有限公司 | Visual detection method and system for stamping processing of automobile parts |
CN116665138B (en) * | 2023-08-01 | 2023-11-07 | 临朐弘泰汽车配件有限公司 | Visual detection method and system for stamping processing of automobile parts |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108986140B (en) | Target scale self-adaptive tracking method based on correlation filtering and color detection | |
CN110353675B (en) | Electroencephalogram signal emotion recognition method and device based on picture generation | |
CN108052980B (en) | Image-based air quality grade detection method | |
KR102284096B1 (en) | System and method for estimating subject image quality using visual saliency and a recording medium having computer readable program for executing the method | |
CN113963193A (en) | Method and device for generating vehicle body color classification model and storage medium | |
CN104346503A (en) | Human face image based emotional health monitoring method and mobile phone | |
CN114998695B (en) | Method and system for improving image recognition speed | |
CN111126481A (en) | Training method and device of neural network model | |
CN110879982A (en) | Crowd counting system and method | |
CN103839042A (en) | Human face recognition method and human face recognition system | |
CN113780106A (en) | Deep learning signal detection method based on radio waveform data input | |
JP2017102622A (en) | Image processing device, image processing method and program | |
CN110704662A (en) | Image classification method and system | |
CN112842348A (en) | Automatic classification method for electrocardiosignals based on feature extraction and deep learning | |
CN110874576A (en) | Pedestrian re-identification method based on canonical correlation analysis fusion features | |
CN107729863B (en) | Human finger vein recognition method | |
CN113486752A (en) | Emotion identification method and system based on electrocardiosignals | |
CN113313179A (en) | Noise image classification method based on l2p norm robust least square method | |
CN109165551B (en) | Expression recognition method for adaptively weighting and fusing significance structure tensor and LBP characteristics | |
CN111488811A (en) | Face recognition method and device, terminal equipment and computer readable medium | |
CN110490210A (en) | A kind of color texture classification method based on compact interchannel t sample differential | |
CN114533084B (en) | Electroencephalogram feature extraction method and device, electronic equipment and storage medium | |
Zhang et al. | A combined approach to single-camera-based lane detection in driverless navigation | |
CN109598262A (en) | A kind of children's facial expression recognizing method | |
CN114912492A (en) | Cross-time individual identification method and system based on electroencephalogram signal deep learning |
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 |