Mobile phone detection door system based on mobile phone component material analysis

Abstract

The invention discloses a mobile phone detection door system based on mobile phone component material analysis, and the system comprises a detection door which is connected with a signal amplification module, the signal amplification module is connected with an analysis processing module, the analysis processing module is connected with an expert database, carries out the comparative analysis of signal data, and is connected with an alarm module, and the alarm module is connected with an alarm module. And giving an alarm for the identified abnormal signal. According to the invention, accurate detection of electronic products such as a mobile phone can be realized, and no matter what state the mobile phone is in, effective identification and alarm can be realized; through a micro residual detection technology and mobile phone component material analysis, the system can eliminate interference of daily carried metal products and reduce the false alarm rate; the system is high in detection speed, is suitable for all confidential places where mobile phones are forbidden to enter, and has important application value; a self-adaptive recognition unit is added, so that characteristic signals similar to materials are recognized, recognition is carried out, and missed judgment is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of mobile phone detection, and in particular to a mobile phone detection gate system based on material analysis of mobile phone components. Background Art

[0002] With the rapid development of information technology, mobile phones have become an indispensable communication tool in people's daily lives. However, in confidential places, carrying and using mobile phones may bring serious security risks, such as information leakage and wiretapping.

[0003] China's authorization publication number CN107238871A discloses a detection system and method for carrying mobile equipment in and out, which uses detection based on micro-residue detection.

[0004] It is designed to detect special objects such as residual materials inside items such as mobile phones. However, based on the current speed of progress in materials science, the existing database may not be able to cope with the ever-changing speed of material iteration, so it is easy to cause misjudgment or omission of judgment due to untimely updates of the internal database, leading to safety hazards. Summary of the invention

[0005] To this end, the present invention provides a mobile phone detection gate system based on the material analysis of mobile phone components to solve the above-mentioned problems in the prior art.

[0006] In order to achieve the above object, the present invention provides the following technical solutions: A mobile phone detection gate system based on mobile phone component material analysis includes: a detection gate connected to a signal amplification module, and the signal amplification module is connected to an analysis and processing module, the analysis and processing module is connected to an expert library, and a comparative analysis is performed on signal data, and the analysis and processing module is connected to an alarm module, and an alarm is issued for an identified abnormal signal, wherein: The detection door contains one or more pairs of electromagnetic induction coils, which emit high-frequency electromagnetic fields and monitor the reflected signals. When someone passes through with an electronic device, the electronic components inside the electronic device interact with the electromagnetic field, generating electromagnetic induction. These induction signals are captured by the receiver in the signal amplification module and converted into recognizable signals. A signal amplification module, used to amplify the identifiable signal converted by the receiver and transmit it to the analysis and processing module; The analysis and processing module captures the amplified signal transmitted by the signal amplification module. The adaptive recognition unit inside it transmits the characteristic information of the amplified signal to the expert database for identification and comparison, and finds the most matching signal feature. After the identification is completed, the information on whether to release the alarm signal is sent to the alarm module.

[0007] Furthermore, the signal amplification module includes a filtering unit, a low-noise amplification unit, and an analog-to-digital converter.

[0008] Furthermore, the expert database stores characteristic information of known signals and their corresponding sources and properties.

[0009] Furthermore, the specific steps of the adaptive recognition unit include: S1: Preprocessing: preprocessing the collected signals to eliminate noise, interference and unnecessary components; S2: Feature extraction, extracting signal features; S3: Intelligent analysis, using convolutional neural networks to further analyze and process the extracted signal features, and compare and judge with the expert database; S4: Result output, outputting the output results of the intelligent analysis to the human-computer interaction interface and alarm module.

[0010] Furthermore, the algorithms used for feature extraction in S2 include: statistical feature extraction, transform domain feature extraction, and time domain feature extraction.

[0011] Furthermore, convolutional neural network is also used for feature extraction in S3.

[0012] Furthermore, the steps of convolutional neural network analysis and processing in S3 include: S3.1: Convert signal features into image data and perform preprocessing; S3.2: Convolutional layer operation: Select a 5x5 convolution kernel, slide the convolution kernel on the input image, and perform a convolution operation on each local area, that is, multiply and sum each element to obtain the value of the corresponding position of the output feature map. After the convolution operation, apply the ReLU function to increase the nonlinear expression ability of the model and suppress the gradient diffusion problem. S3.3: Pooling layer operation: Select average pooling to perform the pooling operation, downsample the feature map output by the convolution layer, and select the maximum value or average value in the local area as the output value of the area; S3.4: Fully connected layer operations: Flatten the feature map output by the pooling layer into a feature vector, input the feature vector into the fully connected layer, perform linear transformation through the weight matrix and bias, and apply the activation function to obtain the final classification result; S3.5: Calculate loss: Select cross entropy loss calculation, calculate the loss function value based on the difference between the output result obtained by forward propagation and the true label, pass the loss function value back to the network, calculate the error layer by layer through the back propagation algorithm, and calculate the gradient of the parameters of each layer, and use the optimization algorithm to update the network parameters according to the gradient; S3.6: Output the result. Input the image to be analyzed into the trained CNN model and obtain the output result through forward propagation.

[0013] Furthermore, the output results include signal recognition results and signal characteristics, as well as signal statistical information and visualization charts.

[0014] The present invention has the following advantages: The present invention can realize accurate detection of electronic products such as mobile phones, and can effectively identify and alarm no matter what state the mobile phone is in; through micro-residue detection technology and mobile phone component material analysis, the system can eliminate the interference of metal products carried daily and reduce the false alarm rate; the system has a fast detection speed and is suitable for all confidential places where mobile phones are prohibited from entering, and has important application value; an adaptive recognition unit is added to identify characteristic signals of similar materials, make identification, and avoid missed judgments. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A module connection diagram of a mobile phone detection gate system based on material analysis of mobile phone components provided by the present invention. DETAILED DESCRIPTION

[0016] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] See also Figure 1 As shown, the present invention provides a mobile phone detection gate system based on the material analysis of mobile phone components, including: a detection gate, which is connected to a signal amplification module, and the signal amplification module is connected to an analysis and processing module, and the analysis and processing module is connected to an expert library to compare and analyze signal data, and the analysis and processing module is connected to an alarm module to issue an alarm for an identified abnormal signal, wherein: The detection gate contains one or more pairs of electromagnetic induction coils, which emit high-frequency electromagnetic fields and monitor the reflected signals. When someone passes through with electronic devices such as mobile phones, the electronic components inside the mobile phone will interact with the electromagnetic field to produce electromagnetic induction. These induction signals are captured by the receiver in the signal amplification module and converted into recognizable signals. Specifically, under certain environmental conditions, items will produce certain residual values ​​of electromagnetic characteristics due to factors such as their material and shape. Such residual values ​​are captured and identified by eddy current classification and identification detection technology and magnetic field conversion technology, thereby obtaining the material and shape information of contraband items. When an item enters the electromagnetic field of a detection device, the metal components or specific materials inside it will interact with the electromagnetic field to produce an eddy current effect, which will be captured by the detection device and converted into a recognizable signal, thereby separating contraband from daily necessities. The sensitivity can detect metal as small as a paper clip, and the sensitivity can be adjusted between 0-100. The possible location, volume, weight, and size of metal objects can be pre-set, and the sensitivity can be adjusted appropriately to eliminate false alarms of items such as belt buckles, keys, jewelry, and coins; The signal amplification module is used to amplify the identifiable signal converted by the receiver and transmit it to the analysis and processing module, which includes a filtering unit, a low-noise amplification unit, and an analog-to-digital converter to perform anti-crosstalk processing, signal-to-noise ratio improvement, digital signal conversion, and signal amplification on the received signal; The analysis and processing module captures the amplified signal transmitted by the signal amplification module. The adaptive recognition unit inside it transmits the characteristic information of the amplified signal, such as frequency, amplitude, phase, etc., to the expert database for identification and comparison, and finds the most matching signal characteristics. The expert database stores a large amount of characteristic information of known signals at home and abroad and their corresponding sources and properties to help identify where the current material source comes from. After the identification is completed, the information on whether to release the alarm signal is sent to the alarm module; The specific steps of the adaptive recognition unit include: S1: Preprocessing: preprocess the collected signals to eliminate noise, interference and unnecessary components. The preprocessing steps include filtering, denoising, standardization, etc. to ensure the quality of the signal and the accuracy of the analysis; S2: Feature extraction, using statistical feature extraction to calculate the statistics of the signal to extract features, such as mean, variance, standard deviation, kurtosis, skewness, etc., which can reflect some basic properties of the signal, such as distribution characteristics, discreteness, etc.; transform domain feature extraction extracts features by converting the signal from the time domain to the frequency domain, time-frequency domain or other transform domains. For example, Fourier transform can convert the signal from the time domain to the frequency domain to extract spectral features, and wavelet transform can provide time-frequency information of the signal at different scales and positions; time domain feature extraction directly analyzes the signal in the time domain to extract time domain features such as zero crossing, peak value, valley value, waveform length, etc. These features can reflect the temporal change law and periodicity of the signal; Convolutional neural networks (CNNs) can also be used to automatically learn and extract high-level features from signals, such as local features, global features, contextual information, etc. S3: Intelligent analysis, using convolutional neural networks to further analyze and process the extracted features, converting the original signal into a high-level feature representation, which is then used for tasks such as signal classification and signal recognition. At the same time, it is compared and judged with the expert database to improve the accuracy and reliability of the analysis; Specifically, the steps of convolutional neural network analysis and processing include: S3.1: Convert signal features into image data and perform preprocessing, including image normalization (scaling pixel values ​​to a suitable range), color space conversion, and data augmentation (such as rotation, scaling, cropping, etc.) to improve the generalization ability and robustness of the model; S3.2: Convolutional layer operation: Select a 5x5 convolution kernel, slide the convolution kernel on the input image, and perform a convolution operation on each local area, that is, multiply and sum each element to obtain the value of the corresponding position of the output feature map. After the convolution operation, apply the ReLU function to increase the nonlinear expression ability of the model and suppress the gradient diffusion problem. S3.3: Pooling layer operation: Select average pooling to perform pooling operations, downsample the feature map output by the convolutional layer, and select the maximum value or average value in a local area as the output value of the area, thereby reducing the dimension and computational complexity of the feature map; S3.4: Fully connected layer operations: Flatten the feature map output by the pooling layer into a feature vector, input the feature vector into the fully connected layer, perform linear transformation through the weight matrix and bias, and apply the activation function to obtain the final classification result or regression value; S3.5: Calculate loss: Select cross entropy loss calculation, calculate the loss function value based on the difference between the output result obtained by forward propagation and the true label, pass the loss function value back to the network, calculate the error layer by layer through the back propagation algorithm, and calculate the gradient of the parameters of each layer, update weights: use optimization algorithms (such as SGD, Adam, etc.) to update network parameters (including convolution kernel weights and biases, etc.) according to the gradient to reduce the loss function value and optimize model performance; S3.6: Output the result. Input the image to be analyzed into the trained CNN model and obtain the output result through forward propagation. S4: Result output: output the results of intelligent analysis to the human-computer interaction interface and alarm module. The output results include signal recognition results, classification labels, intensity levels, etc., as well as related statistical information and visual charts.

[0018] Specifically, objects will produce a special residual value in a certain environment, and then through electromagnetic eddy current detection, adaptive recognition and other technologies, the detected signal is amplified and processed and introduced into the expert database. The data is compared and analyzed to determine whether electronic products such as mobile phones are carried. Micro-residual detection technology is a passive detection method. It will not change the original properties of the detected objects, which means that the technology can be used for detection without destroying the objects, avoiding the damage that may be caused by traditional detection methods, and does not rely on active emission of electromagnetic waves for detection, so it will not produce any radiation effects on the human body. This makes micro-residual detection technology have extremely high application value in confidential places, medical environments and other places that require special protection; The adaptive recognition unit is used to identify the signal of the current detected object, amplify and compare it, and learn at the same time, so as to identify the characteristic signals of similar materials, make identification, avoid missed judgments, and adapt to material iteration; It can also accurately detect electronic products such as mobile phones, and effectively identify and alarm regardless of the state of the mobile phone; Low false alarm rate: Through micro-residue detection technology and mobile phone component material analysis, the system can eliminate the interference of metal products carried daily and reduce the false alarm rate; Fast security check: The system has a fast detection speed and can complete a security check in just 1-2 seconds, which greatly improves the security check efficiency; It is suitable for all confidential places where mobile phones are prohibited, and has important application value.

Claims

1. A mobile phone detection gate system based on material analysis of mobile phone components, characterized in that: include: The detection gate is connected to the signal amplification module, and the signal amplification module is connected to the analysis and processing module, and the analysis and processing module is connected to the expert library to compare and analyze the signal data. The analysis and processing module is connected to the alarm module to issue an alarm for the identified abnormal signal, wherein: The detection door contains one or more pairs of electromagnetic induction coils, which emit high-frequency electromagnetic fields and monitor the reflected signals. When someone passes through with an electronic device, the electronic components inside the electronic device interact with the electromagnetic field, generating electromagnetic induction. These induction signals are captured by the receiver in the signal amplification module and converted into recognizable signals. A signal amplification module, used to amplify the identifiable signal converted by the receiver and transmit it to the analysis and processing module; The analysis and processing module captures the amplified signal transmitted by the signal amplification module. The adaptive recognition unit inside it transmits the characteristic information of the amplified signal to the expert database for identification and comparison, and finds the most matching signal feature. After the identification is completed, the information on whether to release the alarm signal is sent to the alarm module.

2. According to claim 1, a mobile phone detection gate system based on mobile phone component material analysis is characterized in that: The signal amplification module includes a filtering unit, a low-noise amplification unit, and an analog-to-digital converter.

3. According to claim 1, a mobile phone detection gate system based on mobile phone component material analysis is characterized in that: The expert database stores characteristic information of known signals and their corresponding sources and properties.

4. According to claim 1, a mobile phone detection gate system based on mobile phone component material analysis is characterized in that: The specific steps of the adaptive recognition unit include: S1: Preprocessing: preprocessing the collected signals to eliminate noise, interference and unnecessary components; S2: Feature extraction, extracting signal features; S3: Intelligent analysis, using convolutional neural networks to further analyze and process the extracted signal features, and compare and judge with the expert database; S4: Result output, outputting the output results of the intelligent analysis to the human-computer interaction interface and alarm module.

5. A mobile phone detection gate system based on mobile phone component material analysis according to claim 4, characterized in that: The algorithms used for feature extraction in S2 include: statistical feature extraction, transform domain feature extraction, and time domain feature extraction.

6. A mobile phone detection gate system based on mobile phone component material analysis according to claim 4, characterized in that: Convolutional neural network is also used for feature extraction in S3.

7. The mobile phone detection gate system based on mobile phone component material analysis according to claim 4 is characterized in that: The steps of analysis and processing in S3 include: S3.1: Convert signal features into image data and perform preprocessing; S3.2: Convolutional layer operation: Select a 5x5 convolution kernel, slide the convolution kernel on the input image, and perform a convolution operation on each local area, that is, multiply and sum each element to obtain the value of the corresponding position of the output feature map. After the convolution operation, apply the ReLU function to increase the nonlinear expression ability of the model and suppress the gradient diffusion problem. S3.3: Pooling layer operation: Select average pooling to perform the pooling operation, downsample the feature map output by the convolution layer, and select the maximum value or average value in the local area as the output value of the area; S3.4: Fully connected layer operations: Flatten the feature map output by the pooling layer into a feature vector, input the feature vector into the fully connected layer, perform linear transformation through the weight matrix and bias, and apply the activation function to obtain the final classification result; S3.5: Calculate loss: Select cross entropy loss calculation, calculate the loss function value based on the difference between the output result obtained by forward propagation and the true label, pass the loss function value back to the network, calculate the error layer by layer through the back propagation algorithm, and calculate the gradient of the parameters of each layer, and use the optimization algorithm to update the network parameters according to the gradient; S3.6: Output the result. Input the image to be analyzed into the trained CNN model and obtain the output result through forward propagation.

8. A mobile phone detection gate system based on mobile phone component material analysis according to any one of claims 4 to 7, characterized in that: The output results include signal recognition results and signal characteristics, as well as signal statistics and visualization charts.

Images