CN115984285B - Method and system for detecting library bit state based on generation countermeasure network and storage medium - Google Patents
Method and system for detecting library bit state based on generation countermeasure network and storage medium Download PDFInfo
- Publication number
- CN115984285B CN115984285B CN202310277991.7A CN202310277991A CN115984285B CN 115984285 B CN115984285 B CN 115984285B CN 202310277991 A CN202310277991 A CN 202310277991A CN 115984285 B CN115984285 B CN 115984285B
- Authority
- CN
- China
- Prior art keywords
- library
- value
- countermeasure network
- ganomaly
- positive
- 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.)
- Active
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a method and a system for detecting a library bit state based on a generation countermeasure network and a storage medium, wherein the method comprises the following steps: collecting a plurality of pictures in the blank position and the occupied state of the library, and defining the pictures as positive and negative samples respectively; training a GANomaly countermeasure network by taking a positive sample as input until the similarity between the positive sample and a reconstructed graph reaches an expected target, inputting a current true graph of a library into the trained GANomaly countermeasure network, acquiring the reconstructed graph, and performing post-processing to acquire a comparison value; defining a stock state judgment value according to the difference boundary between the corresponding reconstruction images generated in the trained GANomaly countermeasure network by the positive and negative samplesThe method comprises the steps of carrying out a first treatment on the surface of the Compare the comparison value withAnd comparing to judge the corresponding library position state feedback, thereby realizing the judgment of the library position state based on the countermeasure network, reducing the requirements for making training sets and training, and improving the detection adaptability.
Description
Technical Field
The invention relates to an image detection technology, in particular to a detection method and a detection system for detecting whether a library state is empty or not based on a GANomaly generation countermeasure network.
Background
The warehouse position state judgment has wide application value in the field of industrial logistics automation, for example, after whether cargoes exist in a warehouse stacking area or not, the warehouse position state judgment can be conducted by commanding an unmanned forklift through a dispatching system. Therefore, how to accurately judge the change of the warehouse position state is important for automatic warehouse management.
In the prior art, the state of a library is calculated and judged mainly by identifying the position of goods in an image and combining a delimited library region. For example, the prior art provides a library level detection method and system (chinese patent application No. 202210381020.2) wherein the library level detection system comprises at least one camera mounted at a height above a library level area to be detected, the field of view of the at least one camera covering the whole library level area for capturing real-time images of the library level area, the library level area comprising a plurality of library levels, the method comprising: performing stitching processing on the real-time images acquired by at least one camera to obtain a global map of the library position area to be detected; training a target recognition model of a library position area to be detected based on a deep learning framework; and identifying the real-time image through the trained target detection model according to the global map to obtain state information and position information of each library bit in the library bit region.
However, in the concept of the target detection technology adopting the deep learning framework, the probability prediction of whether a library is empty/occupied is obtained, so that during training, sample graphs of different patterns and at different angles of view in two states of the empty library and the occupied library are necessarily required to be simultaneously learned, and therefore, high requirements exist in the aspects of training set production and training.
Secondly, in terms of detection adaptability, in the existing target detection technology, when a new type of goods occupy place on a warehouse location, samples of the new type of goods may need to be collected again to make a training set so as to train an updated model, so that the detection adaptability is weaker.
Disclosure of Invention
Therefore, the main objective of the present invention is to provide a method and a system for detecting a library state based on generation of an countermeasure network, so as to determine the library state based on the countermeasure network, thereby reducing the requirements for making training sets and training, and improving the detection adaptability.
In order to achieve the above object, according to a first aspect of the present invention, there is provided a method for detecting a state of a library based on generation of an countermeasure network, comprising the steps of:
step S100, collecting a plurality of pictures in the blank position and the occupied state of the library, and defining the pictures as positive and negative samples respectively;
step S200, training the GANomaly countermeasure network by taking the positive sample as input until the similarity between the positive sample and the reconstruction thereof is ensured to reach the expected target;
step S300, inputting the current real graph of the library into a trained GANomaly countermeasure network, acquiring a reconstructed graph, and then performing post-processing to acquire a comparison value; defining a stock state judgment value according to the difference boundary between the corresponding reconstruction images generated in the trained GANomaly countermeasure network by the positive and negative samplesThe method comprises the steps of carrying out a first treatment on the surface of the Compare the comparison value with +.>And comparing to judge the corresponding library state feedback.
In a possibly preferred embodiment, in the step S300, the post-processing step of reconstructing includes: performing difference calculation on the obtained corresponding reconstructed image and the real image, performing gray value and self-adaptive threshold binarization processing on the reconstructed image after the difference calculation, performing summation statistics to obtain a comparison value,
in a possibly preferred embodiment, in the step S300: library state judgment value
Wherein->For the maximum difference value between each positive sample and its reconstructed image, +.>For the minimum difference value between each negative sample and its reconstruction, wherein alpha is the degree to which the threshold value is close to the upper and lower boundaries, and the value interval is [0.1,0.9 ]]。
In a possibly preferred embodiment, in the step S300:
wherein the method comprises the steps ofThe method comprises the following steps: positive samples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>M is the number of positive samples; />The method comprises the following steps: negative examples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>N is the number of negative samples; gray is Gray value processing, adaThresh self-adaptive threshold value binarization processing, sum is sum calculation.
In order to achieve the above object, according to a second aspect of the present invention, there is also provided a method for detecting a state of a library based on generation of an countermeasure network, comprising the steps of:
step S100, collecting a plurality of library positions in empty and occupied statesA picture, respectively defined as positive samplexNegative sampley;
Step S200, constructing a GANomaly countermeasure network, and taking positive samplesxAs input, obtain the reconstructed image corresponding to the positive sampleAnd judgexAnd->To generate alternate training against the network until ensuringxAnd->The similarity achieves the expected goal;
step S300, defining a library state judgment valueWherein->For the maximum difference value between each positive sample and its reconstructed image, +.>A minimum difference value between each negative sample and its reconstructed image;
step S400, checking the true map of the library positionx1 inputting the trained GANomaly countermeasure network to obtain a reconstruction graphAnd combine it withx1 calculate the difference->Then, gray value processing and adaptive threshold binarization processing are carried out on the obtained product, and summation statistics is carried out to obtain a comparison valueX2;
Step S500 judgmentX2 is greater thanIf yes, the library bit is in an occupied state, otherwise, the library bit is in a vacant state.
In a possibly preferred embodiment, in the step S300:
wherein the method comprises the steps ofThe method comprises the following steps: positive samples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>M is the number of positive samples; />The method comprises the following steps: negative examples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>N is the number of negative samples; gray is Gray value processing, adaThresh self-adaptive threshold value binarization processing, sum is sum calculation.
In order to achieve the above object, according to a third aspect of the present invention, there is also provided a system for detecting a status of a library based on generation of an countermeasure network, comprising:
a storage unit, configured to store a program including the steps of the method for detecting a status of a library based on generation of an countermeasure network as described in any one of the above, for the control unit, the processing unit, and for timely scheduling execution;
the control unit is used for controlling the camera to collect a plurality of pictures in the empty position and the occupied state of the library, and the pictures are respectively defined as positive and negative samples;
a processing unit for training the GANomaly challenge network with the positive sample as input until ensuring that the positive sample achieves the intended target with its reconstructed similarity;
the control unit is also used for controlling the camera to acquire a current real map of the library position;
the processing unit inputs the real graph into a trained GANomaly countermeasure network to calculate the difference value between the real graph and the reconstructed graph, performs gray value and self-adaptive threshold value binarization processing, and then performs summation statistics to obtain a comparison value; comparing the comparison value with the library state judgment valueAnd comparing to judge the corresponding library state feedback.
To achieve the above object, according to a fourth aspect of the present invention, there is also provided a computer-readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the steps of the method for detecting a status of a library based on generation of an countermeasure network as described in any one of the above.
According to the method and the system for detecting the state of the library based on the generated countermeasure network, the countermeasure network can be trained only through the blank sample of the library, so that compared with the prior art, the difficulty in manufacturing a training set and a training network can be effectively reduced, meanwhile, the inventor skillfully designs that the comparison value is obtained through post-processing of the reconstruction corresponding to the real graph to judge the comparison value with the preset library state judgment value so as to obtain the library state result, once the countermeasure network is trained, the countermeasure network model is not required to be trained and updated no matter what goods occupy the library in the follow-up process, and therefore adaptability is greatly improved, meanwhile, the countermeasure network technology is also skillfully utilized, and the function of detecting the state of the library is realized.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention. In the drawings:
FIG. 1 is a schematic diagram of the steps of a method according to a first embodiment of the present invention;
FIG. 2 is a schematic diagram of a GANomaly challenge network in a method according to a first embodiment of the invention;
FIG. 3 is a schematic diagram of a data processing flow in a method according to a first embodiment of the present invention;
FIG. 4 is an exemplary diagram of positive and negative samples and their corresponding reconstructed and difference binary maps in a method according to a first embodiment of the present invention;
fig. 5 is a schematic diagram of a system structure according to a second embodiment of the present invention.
Detailed Description
In order that those skilled in the art can better understand the technical solutions of the present invention, the following description will clearly and completely describe the specific technical solutions of the present invention in conjunction with the embodiments to help those skilled in the art to further understand the present invention. It will be apparent that the embodiments described herein are merely some, but not all embodiments of the invention. It should be noted that embodiments and features of embodiments in this application may be combined with each other by those of ordinary skill in the art without departing from the inventive concept and conflict. All other embodiments, which are derived from the embodiments herein without creative effort for a person skilled in the art, shall fall within the disclosure and the protection scope of the present invention.
Furthermore, the terms "first," "second," "S100," "S200," and the like in the description and in the claims and drawings are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those described herein. Also, the terms "comprising" and "having" and any variations thereof herein are intended to cover a non-exclusive inclusion. Unless specifically stated or limited otherwise, the terms "disposed," "configured," "mounted," "connected," "coupled" and "connected" are to be construed broadly, e.g., as being either permanently connected, removably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the terms in this case will be understood by those skilled in the art in view of the specific circumstances and in combination with the prior art.
The invention relates to a library position detection, which is used for detecting whether a library position state is abnormal or not, but the problem that the abnormal sample has few samples, more changes and large difference in the abnormal detection process is solved, so that the number of the abnormal sample is too many, but the normal sample (library position empty state sample) is more stable.
For this purpose, as shown in fig. 1 to 4, the method for detecting a status of a library based on a generated countermeasure network according to the present invention includes the steps of:
step S100, positive and negative sample data sets are manufactured: collecting multiple pictures in the blank and occupied state of the library, respectively defining as positive samplesxNegative sampley。
For example, first, in an image including a bin region, positive and negative sample images are boxed and cut out as shown in fig. 4, wherein the number of positive sample training samples is preferably S, the number of positive sample test samples is 0.2×s, and the number of negative sample test samples is 0.2×s.
Secondly, positive and negative samples are takenx、yThe training set contains only positive samples, uniformly scaled to a fixed size, e.g. (64 x 64).
Step S200, constructing a GANomaly countermeasure network, and taking positive samplesxAs input, obtain the reconstructed image corresponding to the positive sampleAnd judgexAnd->To generate a countering network alternate training until the expected goal is reached, ensurexAnd->The similarity is as high as possible.
For example, as shown in fig. 2, the overall structure of the GANomaly is that the upper half adopts a network structure of an encoder, a decoder and an encoder, and learns the original image to generate a reconstruction, the latent expression of the original image generates the latent expression of the reconstruction, and the lower half defines a discriminator. The network and the countermeasure network are generated through alternate training.
Wherein the generator networkG: is composed of two parts, encoderAnd decoder->. Belonging to self-coding structure for learning input +.>Is to reconstruct an image +.>。/>Input of +.>Dimension->Generating a vector via a network>Its dimension is。/>Is to add vector->Conversion to->Its dimension and->And consistent. />The transposed convolution is mainly applied.
Encoder networkE: encoder E structureIs the same, pseudo-image to be generated +.>Turn to->The encoder network is used for replacing the process of searching the potential expression of the pseudo image in the testing stage in the AnoGan design (the potential expression is the final obtained characteristic of the conventional convolution network), so that the efficiency is improved.
Distinguishing device networkD: distinguishing device networkThere are two inputs, original diagram +.>Restructuring +_>. And finally, calculating the final score through convolution, and judging the true or false.
Loss function: the loss function of the GANomaly network has an upper generator loss function and a lower arbiter loss function. The upper part has three parts:。
wherein the method comprises the steps ofFor the original image +.>Is->Distribution of->Is the intermediate layer output of the arbiter network, +.>The output of (2) is +.>。。
The total loss function of the generator is:
The Loss of the arbiter is consistent with the original GAN, BCE Loss:
wherein the method comprises the steps ofIndicate probability distribution obeyed by true graph x, +.>Refers to the result of the arbiter network.
During network training, the generator network of the upper part and the discriminator network of the lower part are trained alternately, and the training process comprises the following steps:
step S201 is to train the generator and the arbiter of GANomaly in a cyclic and alternating mode, wherein the gradient updating mode is Adam, and the initial value of the learning rate is 0.0002. An example of a data batch size is 64.
Step S202, loop training completion conditions: difference between positive and reconstructed samples in a test sample setMaximum valueLess than the minimum value of the difference between the negative and reconstructed samples +.>The method comprises the steps of carrying out a first treatment on the surface of the The number of training alternations reaches a set number of cycles. Otherwise readjusting the sample set.
After the training of step S203 is completed, calculationAs the judgment basis for the subsequent reasoning.
Step S300 calculates a library status determination valueWherein->For the maximum difference value between each positive sample and its reconstructed image, +.>For the minimum difference value between each negative sample and its reconstruction, wherein alpha is the degree to which the threshold value is close to the upper and lower boundaries, the value interval is preferably [0.1,0.9 ]]。
Specifically, the present invention uses GANomaly for bin detection, not by the true-false score of the arbiter. The role of the discriminator is to ensure the original positive sample imageIs recombined with->The similarity of (c) is as high as possible. The real samples of the training contain only positive samples, so the reconstructed image of the similar positive samples (empty bin) differs less from the original image, while the reconstructed image of the negative samples (non-empty bin) differs more from the original negative sample.
The library bit detection logic of the scheme is to compare the difference value between the normal sample and the reconstructed sampleNegative sample of test->Is recombined with->Difference of->Difference between them. And then respectively carrying out grey value treatment on the obtained products, carrying out self-adaptive threshold value binarization treatment, and carrying out summation statistics.
For example:
wherein the method comprises the steps ofThe method comprises the following steps: positive samples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>M is the positive sample number.
The method comprises the following steps: negative examples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>N is the number of negative samples; gray is Gray value processing, adaThresh self-adaptive threshold value binarization processing, sum is sum calculation.
Further, in the preferred embodiment, the inventors have adjusted through a number of experiments, preferablyWhen the optimal selection value is 0.2, the method can better reflect +.>To improve the subsequent judgment +.>To identify the accuracy of the bin status.
Step S400, true images of the library positions to be checkedx1 inputting the trained GANomaly countermeasure network to obtain a reconstruction graphAnd combine it withx1 calculate the difference->Then, gray value processing and adaptive threshold binarization processing are carried out on the obtained product, and summation statistics is carried out to obtain a comparison valueX2。
Specifically, after training the GANomaly challenge network, and computing to obtainThen, the formal state detection of the library can be started, wherein the real image of the currently detected library is preprocessed as the positive sample, and then used as the input of the trained GANomaly countermeasure network, so as to obtain the corresponding reconstructed image and then perform post-processing (the post-processing preferably adopts an OpenCV function library) for calculating the comparison valueX2:
That is, the obtained corresponding reconstruction image and the real image are subjected to difference calculation, the reconstruction image after the difference calculation is subjected to gray value and adaptive threshold binarization, and then summation statistics is carried out, so as to obtain a comparison value, whereinRepresenting the absolute value function->Represents a gray-scale function, ">Representing an adaptive threshold binarization function, sum representing a summation function.
Step S500 finally, the comparison value obtained in step S400X2 judging whether or notX2>If yes, the library bit is in an occupied state, otherwise, the library bit is in a vacant state.
Thus, unlike the conventional GANomaly network, the conventional GANomaly network distinguishes positive and negative samples by determining the score of the network, and in this scheme, the score of the determination network is not referred to, but the difference between the result of the generation network and the original input is calculated. The method is used for detecting the library position by using a sample with only empty library position in training, has good generating effect on the sample, but has poor generating effect on non-empty library bitmap, and has larger difference between an input picture and a generating result under two conditions, and exactly corresponds to the conception of two states of occupied and unoccupied library position, thereby realizing the detection of the library position state by skillfully using the characteristics of a GANomaly network.
In accordance with the first embodiment, referring to fig. 5, the present invention further provides a system for detecting a status of a library based on generating an countermeasure network, which includes:
a storage unit, configured to store a program including steps corresponding to the method for detecting a status of a library based on generation of an countermeasure network as shown in the first embodiment, for the control unit, the processing unit, and the timely fetch and execute the program;
the control unit is used for controlling the camera to collect a plurality of pictures in the empty position and the occupied state of the library, and the pictures are respectively defined as positive and negative samples;
the processing unit is used for taking the positive sample as input to train the GANomaly countermeasure network until the expected target is achieved, and ensuring that the similarity between the positive sample and the reconstruction graph is as high as possible;
the control unit is further used for controlling the camera to acquire a current real map of the library position;
the processing unit further inputs the real graph into a trained GANomaly countermeasure network to calculate the difference value between the real graph and the reconstructed graph, performs gray value and self-adaptive threshold value binarization processing, and then performs summation statistics to obtain a comparison value; comparing the comparison value with the library state judgment valueAnd comparing to judge the corresponding library state feedback.
The present invention also provides a computer readable storage medium having a computer program stored thereon, wherein the computer program when executed by a processor implements the steps of the method for detecting a status of a library based on generating an countermeasure network described in the first embodiment.
In summary, according to the method and the system for detecting the state of the library based on the generated countermeasure network, the countermeasure network can be trained only by the blank sample of the library, so that compared with the prior art, the difficulty in manufacturing a training set and a training network can be effectively reduced, meanwhile, the inventor skillfully designs that the comparison value is obtained by performing post-processing on the reconstruction corresponding to the true graph to judge the comparison value with the preset library state judgment value so as to obtain the library state result, and therefore, once the countermeasure network training is completed, no matter what goods occupy the library, the countermeasure network model is not required to be trained and updated any more, thereby greatly improving the adaptability, and meanwhile, the countermeasure network technology is also skillfully utilized, so that the function of detecting the state of the library is realized.
The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is to be limited only by the following claims and their full scope and equivalents, and any modifications, equivalents, improvements, etc., which fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
It will be appreciated by those skilled in the art that the system, apparatus and their respective modules provided by the present invention may be implemented entirely by logic programming method steps, in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc., except for implementing the system, apparatus and their respective modules provided by the present invention in a purely computer readable program code. Therefore, the system, the apparatus, and the respective modules thereof provided by the present invention may be regarded as one hardware component, and the modules included therein for implementing various programs may also be regarded as structures within the hardware component; modules for implementing various functions may also be regarded as being either software programs for implementing the methods or structures within hardware components.
Furthermore, all or part of the steps in implementing the methods of the embodiments described above may be implemented by a program, where the program is stored in a storage medium and includes several instructions for causing a single-chip microcomputer, chip or processor (processor) to perform all or part of the steps in the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In addition, any combination of various embodiments of the present invention may be performed, so long as the concept of the embodiments of the present invention is not violated, and the disclosure of the embodiments of the present invention should also be considered.
Claims (7)
1. The method for detecting the state of the library based on the generation countermeasure network is characterized by comprising the following steps:
step S100, collecting a plurality of pictures in the blank position and the occupied state of the library, and defining the pictures as positive and negative samples respectively;
step S200, training the GANomaly countermeasure network by taking the positive sample as input until the similarity between the positive sample and the reconstruction thereof is ensured to reach the expected target;
step S300, inputting the current real graph of the library into a trained GANomaly countermeasure network, acquiring a reconstructed graph, and then performing post-processing to acquire a comparison value; defining a stock state judgment value according to the difference boundary between the corresponding reconstruction images generated in the trained GANomaly countermeasure network by the positive and negative samplesThe method comprises the steps of carrying out a first treatment on the surface of the Compare the comparison value with +.>Comparing to determine corresponding library state feedback, wherein the library state determination value +.>Wherein->For the maximum difference value between each positive sample and its reconstructed image, +.>The minimum difference value between each negative sample and its reconstructed image is taken, where alpha is the extent to which the threshold is close to the upper and lower boundaries.
2. The method for detecting a status of a library based on generation of an countermeasure network according to claim 1, wherein in the step S300, the post-processing step of the reconstruction includes: and carrying out difference calculation on the obtained corresponding reconstructed image and the real image, carrying out gray value and self-adaptive threshold binarization processing on the reconstructed image after the difference calculation, and then carrying out summation statistics to obtain a comparison value.
3. The method for detecting a status of a library based on generation of an countermeasure network according to claim 1, wherein in the step S300:
wherein the method comprises the steps ofThe method comprises the following steps: positive samples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>M is the number of positive samples; />The method comprises the following steps: negative examples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>N is the number of negative samples; gray is Gray value processing, adaThresh self-adaptive threshold value binarization processing, sum is sum calculation.
4. The method for detecting the state of the library based on the generation countermeasure network is characterized by comprising the following steps:
step S100, collecting multiple pictures in the empty and occupied state of the library, respectively defining as positive samplesxNegative sampley;
Step S200, constructing a GANomaly countermeasure network, and taking positive samplesxAs input, obtain the reconstructed image corresponding to the positive sampleAnd judgexAnd->To generate alternate training against the network until ensuringxAnd->The similarity achieves the expected goal;
step S300, defining a library state judgment valueWherein->For the maximum difference value between each positive sample and its reconstructed image, +.>A minimum difference value between each negative sample and its reconstructed image;
step S400, the true diagram of the library position to be checkedx1 inputting the trained GANomaly countermeasure network to obtain a reconstruction graphAnd will->And (3) withx1 difference calculation +.>Then, after gray value treatment and adaptive threshold value binarization treatment are carried out on the x1', summation statistics is carried out to obtain a comparison valueX2;
5. The method for detecting a status of a library based on generation of an countermeasure network according to claim 4, wherein in the step S300:
wherein the method comprises the steps ofThe method comprises the following steps: positive samples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>M is the number of positive samples; />The method comprises the following steps: negative examples->Inputting trained GANomaly challenge network to obtain its reconstruction +.>To be in charge of->Calculating absolute value of difference>N is the number of negative samples; gray is Gray value processing, adaThresh self-adaptive threshold value binarization processing, sum is sum calculation.
6. A system for detecting a status of a library based on generating an countermeasure network, comprising:
a storage unit for storing a program including the steps of the method for detecting a status of a library based on generation of an countermeasure network according to any one of claims 1 to 5, for the control unit, the processing unit, and for timely scheduling execution;
the control unit is used for controlling the camera to collect a plurality of pictures in the empty position and the occupied state of the library, and the pictures are respectively defined as positive and negative samples;
a processing unit for training the GANomaly challenge network with the positive sample as input until ensuring that the positive sample achieves the intended target with its reconstructed similarity;
the control unit is also used for controlling the camera to acquire a current real map of the library position;
the processing unit inputs the real graph into a trained GANomaly countermeasure network to calculate the difference value between the real graph and the reconstructed graph, performs gray value and self-adaptive threshold value binarization processing, and then performs summation statistics to obtain a comparison value; comparing the comparison value with the library state judgment valueAnd comparing to judge the corresponding library state feedback.
7. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310277991.7A CN115984285B (en) | 2023-03-21 | 2023-03-21 | Method and system for detecting library bit state based on generation countermeasure network and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310277991.7A CN115984285B (en) | 2023-03-21 | 2023-03-21 | Method and system for detecting library bit state based on generation countermeasure network and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115984285A CN115984285A (en) | 2023-04-18 |
CN115984285B true CN115984285B (en) | 2023-07-04 |
Family
ID=85960003
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310277991.7A Active CN115984285B (en) | 2023-03-21 | 2023-03-21 | Method and system for detecting library bit state based on generation countermeasure network and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115984285B (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111598877A (en) * | 2020-05-18 | 2020-08-28 | 河北工业大学 | Lithium battery surface defect detection method based on generation of countermeasure network |
CN113688889A (en) * | 2021-08-13 | 2021-11-23 | 上海商汤智能科技有限公司 | Abnormality detection method, abnormality detection device, electronic apparatus, and computer-readable storage medium |
Family Cites Families (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109584221B (en) * | 2018-11-16 | 2020-07-28 | 聚时科技(上海)有限公司 | Abnormal image detection method based on supervised generation countermeasure network |
CN110245619B (en) * | 2019-06-17 | 2020-11-27 | 杭州海康睿和物联网技术有限公司 | Method and system for judging overrun object on escalator |
CN110781780B (en) * | 2019-10-11 | 2023-04-07 | 浙江大华技术股份有限公司 | Vacancy detection method and related device |
CN114746908A (en) * | 2020-01-23 | 2022-07-12 | 瑞典爱立信有限公司 | Method and apparatus for image classification |
CN112270298B (en) * | 2020-11-16 | 2023-04-25 | 北京深睿博联科技有限责任公司 | Method, device and equipment for identifying road abnormality and computer readable storage medium |
CN113688857A (en) * | 2021-04-26 | 2021-11-23 | 贵州电网有限责任公司 | Method for detecting foreign matters in power inspection image based on generation countermeasure network |
CN114627468A (en) * | 2022-03-16 | 2022-06-14 | 青岛理工大学 | Wheat field obstacle detection method based on generation of confrontation network |
-
2023
- 2023-03-21 CN CN202310277991.7A patent/CN115984285B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111598877A (en) * | 2020-05-18 | 2020-08-28 | 河北工业大学 | Lithium battery surface defect detection method based on generation of countermeasure network |
CN113688889A (en) * | 2021-08-13 | 2021-11-23 | 上海商汤智能科技有限公司 | Abnormality detection method, abnormality detection device, electronic apparatus, and computer-readable storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN115984285A (en) | 2023-04-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110135503B (en) | Deep learning identification method for parts of assembly robot | |
CN111814852B (en) | Image detection method, image detection device, electronic equipment and computer readable storage medium | |
Fioravanti et al. | A deep artificial immune system to detect weld defects in DWDI radiographic images of petroleum pipes | |
CN116740728B (en) | Dynamic acquisition method and system for wafer code reader | |
US20230201973A1 (en) | System and method for automatic detection of welding tasks | |
CN111178446A (en) | Target classification model optimization method and device based on neural network | |
CN115982235A (en) | Abnormal time sequence data detection method, equipment and medium | |
JP2021174456A (en) | Abnormality determination method and abnormality determination device | |
CN111950633A (en) | Neural network training method, neural network target detection method, neural network training device, neural network target detection device and storage medium | |
CN113986561A (en) | Artificial intelligence task processing method and device, electronic equipment and readable storage medium | |
KR20200063361A (en) | A feature level fusion system and method for stock price forecasting | |
CN115909157A (en) | Machine vision-based identification detection method, device, equipment and medium | |
CN115793590A (en) | Data processing method and platform suitable for system safety operation and maintenance | |
CN117474918B (en) | Abnormality detection method and device, electronic device, and storage medium | |
CN116992953B (en) | Model training method, fault diagnosis method and device | |
CN102713974A (en) | Learning device, identification device, learning identification system and learning identification device | |
CN115984285B (en) | Method and system for detecting library bit state based on generation countermeasure network and storage medium | |
CN111368865A (en) | Method and device for detecting remote sensing image oil storage tank, readable storage medium and equipment | |
Liu et al. | A dual-branch balance saliency model based on discriminative feature for fabric defect detection | |
CN113392455A (en) | House type graph scale detection method and device based on deep learning and electronic equipment | |
CN111353577B (en) | Multi-task-based cascade combination model optimization method and device and terminal equipment | |
CN116467652A (en) | Bearing fault prediction method and system based on convolution noise reduction self-encoder | |
CN115984671A (en) | Model online updating method and device, electronic equipment and readable storage medium | |
CN113807541B (en) | Fairness repair method, system, equipment and storage medium for decision system | |
CN115359484A (en) | Image processing method, device, equipment and storage medium |
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 | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A method and system for detecting storage location status based on generative adversarial networks Effective date of registration: 20230828 Granted publication date: 20230704 Pledgee: Bank of Communications Ltd. Shanghai New District Branch Pledgor: Shanghai Xiangong Intelligent Technology Co.,Ltd. Registration number: Y2023310000491 |