Intelligent detection method and system for metal detector

Abstract

The present invention relates to an intelligent detection method and system for a metal detector. The method comprises: preparing pre-recorded images of products, wherein the pre-recorded images at least comprise images of qualified products; acquiring feature vectors of a product at a plurality of frequencies; on the basis of the feature vectors, obtaining feature images; and a metal detector processing the feature images, then comparing same with the pre-recorded images, and outputting a detection determination result of the product. Compared with the prior art, the present invention has the advantages of realizing the detection of non-metallic foreign objects and small-sized metals, achieving high detection accuracy, implementing intelligent detection on the basis of feature images, etc.

Description

Intelligent detection method and system of metal detector TECHNICAL FIELD

[0001] The present application relates to the technical field of detection, in particular to an intelligent detection method and system of a metal detector. BACKGROUND

[0002] A metal detector is an instrument for detecting metal foreign matter in a measured object. The metal detector detects foreign matter, especially metal foreign matter, according to the change of the electromagnetic field in the probe window. The product passing through the probe also has conductivity due to the moisture and salt contained therein, which can significantly change the electromagnetic field in the probe (so-called product effect), and the signal generated thereby can sometimes drown out the metal foreign matter signal. An important task of the metal detector is to distinguish the signal of the metal foreign matter from the effect caused by the product and to remove the product containing the foreign matter.

[0003] The metal detector can currently only detect metal foreign matter, but the foreign matter mixed into the actual product is far more than metal. Common non-metal foreign matter includes stones, glass, soil, paper, plastic products, etc., which cannot be detected by the current scheme. However, these foreign matters mixed into the product will also bring risks.

[0004] Currently, the common means for detecting non-metal foreign matter is to use an X-ray machine, a visual detection device, etc., but each has its own limitations. For example, the X-ray machine is limited to some high-density foreign matter (such as stones, glass, etc.), and it is difficult to act on low-density foreign matter. The visual detection device can detect various foreign matter on the surface of the product, but it cannot detect foreign matter inside the product. Moreover, the cost of these devices for detecting by spectrum is generally high at the present stage.

[0005] How to improve the detection accuracy becomes a technical problem to be solved. SUMMARY

[0006] The purpose of the present application is to overcome the defects of the prior art and provide an intelligent detection method and system of a metal detector.

[0007] The purpose of the present application can be achieved by the following technical solutions.

[0008] According to one aspect of the present application, an intelligent detection method of a metal detector is provided, which comprises:

[0009] 1) preparing a pre-recorded image of the product, including at least an image of a qualified product;

[0010] 2) obtaining a feature vector of the product at multiple frequencies;

[0011] 3) obtaining a feature image based on the feature vector of step 2);

[0012] 4) the metal detector compares the feature image with the pre-recorded image of step 1) after processing the feature image, and outputs the detection result of the product.

[0013] Preferably, the feature vectors of the plurality of frequencies are obtained by adjusting the phase of the carrier signal or the phase of the transmitted signal.

[0014] The feature vector formed by the two signals is expressed as a function of time t:

[0015]

[0016] wherein, are two unit vectors of a two-dimensional coordinate system, and represent the change information of the disturbance generated by the probe over a period of time.

[0017] Preferably, the product feature image is a binary closed curve image formed by the end curve of the feature vector.

[0018] Preferably, the feature images of the plurality of frequencies are subjected to correlation operations, and the results of the correlation operations form new feature images, wherein the correlation operations include difference operations, addition operations, multiplication operations, and fusion operations.

[0019] Preferably, the processing of the feature image by the metal detector includes image preprocessing, image feature extraction, and target recognition detection; the metal detector compares the feature image with the image of the qualified product, and outputs the detection result of the product according to the similarity.

[0020] More preferably, the detection result of the metal detector includes detection of qualified products, unqualified products containing non-metallic foreign matter, unqualified products containing metallic foreign matter, missing parts, and multiple parts.

[0021] Preferably, the pre-recorded image further includes images of unqualified products containing non-metallic foreign matter and images of unqualified products containing metallic foreign matter.

[0022] Preferably, the intelligent detection of the metal detector is used in parallel with the traditional threshold detection.

[0023] According to another aspect of the present application, an intelligent detection system of a metal detector is provided, which includes an actuator, and a probe, a balanced signal generation module, a signal processing circuit, and an AD sampling module connected in sequence, and further includes a signal processing circuit, an intelligent detection module, and a judgment module.

[0024] The intelligent detection module generates a feature image based on the feature vector output by the AD sampling module, and processes the feature image to distinguish the difference from the standard qualified product feature image, wherein the processing includes image preprocessing, image feature extraction and target recognition detection; the judgment module judges the output result of the intelligent detection module, outputs a detection judgment result, and then notifies the execution mechanism to make a corresponding response action.

[0025] Preferably, the system further comprises a threshold detection module parallel to the intelligent detection module, which takes a preset threshold as a judgment standard, and if the sampling signal exceeds the threshold, it is determined that a foreign matter is detected.

[0026] Compared with the prior art, the present application has the following beneficial effects:

[0027] 1) The present application obtains more feature information by obtaining feature vectors of multiple frequencies of the product and obtaining a feature image based on the feature vectors, and improves the detection accuracy by using image processing methods.

[0028] 2) When the pre-recorded image of the present application contains the image of a non-metal foreign matter, the non-metal foreign matter can be detected from the product, solving the previously unsolvable detection problem of non-metallic abnormalities.

[0029] 3) The present application and the traditional threshold detection are parallel, which further improves the detection accuracy.

[0030] 4) Based on the difference of the feature image, the metal detector can detect smaller metal foreign matters according to the image features and using image processing algorithms, improve the detection accuracy and adaptability, and realize intelligent detection of various foreign matters of the product.

[0031] 5) The present application widens the application range of detection and can be used for product foreign matter detection, product missing parts and multiple part detection. BRIEF DESCRIPTION OF DRAWINGS

[0032] Fig. 1 is a schematic diagram of the acquisition process of the balanced signal in the present application;

[0033] Fig. 2 is a schematic diagram of the processing method of the balanced signal in the present application;

[0034] Fig. 3 is a schematic diagram of the feature vector in the present application;

[0035] Fig. 4 is a schematic diagram of the feature image formed by the end curve of the feature vector in the present application;

[0036] Fig. 5 is a schematic diagram of a part of the feature image exceeding the threshold rectangular window in the present application;

[0037] Fig. 6 is a schematic diagram of the feature image being entirely within the threshold rectangular window in the present application;

[0038] Fig. 7 is a schematic diagram of a characteristic image of the first frequency in the present application;

[0039] Fig. 8 is a schematic diagram of a characteristic image of the second frequency in the present application;

[0040] Fig. 9 is a schematic diagram of a characteristic image of a qualified product in the present application;

[0041] Fig. 10 is a schematic diagram of a characteristic image containing non-metals in the present application;

[0042] Fig. 11 is a schematic diagram of a characteristic image containing metals in the present application;

[0043] Fig. 12 is a schematic diagram of a characteristic image of a missing part in the present application;

[0044] Fig. 13 is a schematic diagram of a characteristic image of multiple parts in the present application;

[0045] Fig. 14 is a schematic diagram of a parallel detection system of threshold detection and intelligent detection in the present application;

[0046] Fig. 15 is a schematic diagram of an intelligent detection process in the present application. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts should fall within the scope of protection of the present application.

[0048] The present embodiment relates to an intelligent detection method of a metal detector.

[0049] The metal detector realizes the detection of metal foreign matters by processing the balance signal. The balance signal can be obtained in the following manner. The transmitting signal is amplified by a power amplifier circuit, sent into a transmitting coil through a transmitting transformer, and generates an electromagnetic field of a certain intensity in the probe. The receiving coil senses the change of the electromagnetic field, and its output is sent into a receiving amplifier circuit through a receiving transformer, and outputs the balance signal, as shown in Fig. 1.

[0050] For balanced signals, common processing methods are shown in Figure 2 (including but not limited to). The balanced signal and the quadrature carrier signal generated by the carrier signal generator are input into a multiplier. The output signal of the multiplier is filtered and amplified to obtain the DT1 and DT2 signals used for sampling. After sampling by an AD converter, they are sent to the CPU. Since the balanced signal and the quadrature carrier signal generated by the carrier signal generator are from different sources, the phase of the balanced signal can be changed by adjusting the phase of the transmitted signal (while the phase of the carrier signal remains fixed), or conversely, the phase of the transmitted signal can be fixed while the phase of the carrier signal is adjusted. DT1 and DT2 are processed in the CPU, generally using a preset threshold as the judgment standard. If DT1 or DT2 exceeds the threshold, it is determined that a foreign object has been detected.

[0051] The essence of the above scheme is to obtain, through certain means, the quantitative information of two characteristics of the disturbance generated when the product passes through the probe at a certain time t. and Then, this quantified information is processed using certain judgment rules (thresholds THS1 and THS2) to obtain the detection result.

[0052] It is evident that this processing method loses information about the changes in two characteristics of the disturbance generated by the probe over a period of time, namely ( and ).

[0053] Characterization of a specific product (when passing through the probe) and Together, they reflect the characteristics of this product and can be represented by feature vectors. The representation is shown in Figure 3. Based on the aforementioned method for obtaining DT1 and DT2, it can be seen that DT1 and DT2 are relatively independent. Therefore, this feature vector... The variation with t can be expressed as:

[0054]

[0055] in, These are two unit vectors in a two-dimensional coordinate system.

[0056] The aforementioned method for processing the detection signal (exceeding the threshold) uses a rectangular window to determine this. For a period of time The feature image formed by the vector curve in the above two-dimensional coordinate system The judgment is made as shown in Figure 4. If the feature image contains a portion outside the window, it is considered that a foreign object has been detected, as shown in Figures 5 and 6. In other words, this processing method only considers the relationship between the feature image and the window boundary, losing the information contained in the image itself. However, the information contained in the image itself has more value and can reflect more (more complete) feature information of the product. This is also the advantage of this invention; more complete feature information means better detection results.

[0057] In fact, the two orthogonal signals generated by the carrier signal generator in the above metal detector circuit scheme can also generate two signals with a 120° phase difference, or even add another signal to form three signals with a 120° phase difference between each pair. For example:

[0058]

[0059] in, These are three unit vectors in a two-dimensional coordinate system, and z(t) is the third signal.

[0060] Theoretically, the number of constituent vectors can be increased infinitely. This allows us to obtain more feature information by using the components of the image. However, generally speaking, using a two-dimensional coordinate system to acquire images is a better choice for easily applying mature image processing algorithms.

[0061] In addition, dual-frequency technology can be used to obtain product characteristics on another frequency. Thus, the image formed by its vector curve is obtained. At this point, we have obtained two images that reflect the product's characteristics. and As shown in Figures 7 and 8, these images can be easily processed using various image processing methods, enabling the differentiation of certain non-metallic foreign objects and smaller metallic foreign objects from the product. Of course, this can be extended to more frequencies, resulting in more images. , etc).

[0062] The above describes how a metal detector acquires feature images of objects passing through its probe. The method. Below we will use... Let's take an example to discuss this further.

[0063] Because the receiving coil of the probe is symmetrical, the image of the object passing through the probe is also symmetrical (with appropriate circuit gain). For simplicity, we will only look at... Half of the content.

[0064] Figures 9, 10, and 11 show qualified products without foreign objects. , unqualified products containing non-metallic foreign matter (stones) , unqualified products containing metallic foreign matter .

[0065] Prepare a pre-recorded image of the product, including at least the image of a qualified product;

[0066] Pre-record the image of the qualified product for detection determination by the metal detector. In addition, the pre-recorded image also includes the image of unqualified products containing non-metallic foreign matter, and the image of unqualified products containing metallic foreign matter, for further refinement of the detection result.

[0067] In addition, for application scenarios with missing or extra piece detection requirements, such as a 12-piece cake box that needs to detect whether there is one less, or a case of mineral water that needs to detect whether there is one less, or a bag of 20 chicken legs that needs to detect whether there is one more or one less, the conventional method is to use a checkweigher to judge by weight. The scheme in the present application can realize missing piece ( ) and extra piece ( ) detection, as shown in Figures 12 and 13, respectively.

[0068] If the threshold detection method is used to judge, there may be a case of missing (or extra) pieces but containing foreign matter, which may cause the signal to still be within the threshold range. However, the image intelligent processing method can effectively avoid this situation, because the image characteristics of missing / extra pieces and the image characteristics of containing foreign matter are significantly different.

[0069] The above , , , , cover most of the product feature images, so next we only need to input the image information into the intelligent detection module, process it using various image algorithms, and distinguish the differences from the standard qualified product feature image, to identify various situations, so as to realize the final result determination and notify the execution mechanism to take corresponding response actions. This detection can be parallel to the traditional threshold detection method, and the two are not contradictory, as shown in Figure 14.

[0070] As mentioned above: product feature images at different frequencies and (and even the third and fourth frequencies) contain more information. Therefore, they can be sent to the intelligent detection module for parallel processing. Further, even and can be difference operated, and the result of the difference operation It can also be regarded as a product feature image. Of course, these operations all need to consume more computing resources, and the detection effect and computing efficiency should be balanced in practical application. In addition, in addition to difference operation, other related operations can also be performed, wherein the related operations include but are not limited to addition operation, multiplication operation and fusion operation.

[0071] Generally, as shown in FIG. 15, after the product image is sent into the intelligent detection module, it will go through image preprocessing, image feature extraction, target detection and recognition, and then output the judgment result. In these links, there are a large number of algorithms to choose from for each link, and we need to select the appropriate algorithm according to the image features.

[0072] Since is a binary closed curve image, some suitable algorithms include but are not limited to image stretching, gradient sharpening, Roberts operator, image reconstruction, geometric transformation, image cropping, feature point detection, feature point matching, CNN deep learning, etc.

[0073] These algorithms can be easily obtained through MATLAB, OPENCV and other tools and applied to the device. Especially in the feature extraction and target recognition link, OPENAI can also be used to improve the accuracy and adaptability of processing, and truly realize intelligent detection.

[0074] The embodiment also relates to an intelligent detection system of a metal detector, which comprises a probe, a balance signal generation module, a signal processing circuit and an AD sampling module, a CPU module, a judgment module and an execution mechanism connected in sequence. The CPU module comprises an intelligent detection module and a judgment module.

[0075] The intelligent detection module generates a feature image based on the feature vector output by the AD sampling module, and processes the feature image to distinguish the difference from the standard qualified product image, wherein the processing includes image preprocessing, image feature extraction and target recognition detection.

[0076] The judgment module judges the output result of the intelligent detection module, outputs the detection judgment result, and then notifies the execution mechanism to make a corresponding response action.

[0077] The system also comprises a threshold detection module parallel to the intelligent detection module, and the threshold detection module takes a preset threshold as a judgment standard. If the sampling signal exceeds the threshold, it is judged that a foreign matter is detected. The results of the two detection modules are sent to the judgment module for detection result judgment.

[0078] The above merely illustrates the specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any skilled person in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements shall be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for intelligent detection of a metal detector, characterized in that, The method comprises: 1) preparing pre-recorded images of products, at least including images of qualified products; 2) obtaining feature vectors of the products at multiple frequencies; 3) obtaining product feature images based on the feature vectors of step 2); 4) comparing the product feature images processed by the metal detector with the pre-recorded images of step 1), and outputting the detection results of the products.

2. The intelligent detection method of the metal detector according to claim 1, characterized in that, The feature vectors at multiple frequencies are obtained by adjusting the phase of the carrier signal or the phase of the transmitted signal. The feature vector formed from the two signals The change over time t is represented as: wherein, are two unit vectors of a two-dimensional coordinate system, and The change information of the representation of the disturbance generated by the probe over a period of time.

3. The intelligent detection method of the metal detector according to claim 1, wherein, The product feature image is a binary closed curve image formed based on the feature vector.

4. The intelligent detection method of the metal detector according to claim 1, wherein, The results of the correlation operation of the feature images at multiple frequencies form new feature images, wherein the correlation operation includes difference operation, addition operation, multiplication operation and fusion operation.

5. The intelligent detection method of the metal detector according to claim 1, wherein, The processing of the product feature images by the metal detector includes image preprocessing, image feature extraction and target recognition detection; the metal detector compares the product feature images with the images of qualified products, and outputs the detection results of the products according to the similarity.

6. The intelligent detection method of the metal detector according to claim 5, wherein, The detection results of the metal detector include qualified products, unqualified products containing non-metallic foreign matter, unqualified products containing metallic foreign matter, missing parts and multiple parts.

7. The intelligent detection method of the metal detector according to claim 1, wherein, The pre-recorded images also include images of unqualified products containing non-metallic foreign matter and images of unqualified products containing metallic foreign matter.

8. The intelligent detection method of the metal detector according to claim 1, wherein, The intelligent detection of the metal detector is used in parallel with the traditional threshold detection.

9. A system of intelligent detection method using the metal detector of claim 1, the system comprising an actuating mechanism and sequentially connected probe, balanced signal generation module, signal processing circuit and AD sampling module, characterized in that, The system further comprises a signal processing circuit, an intelligent detection module and a judgment module; The intelligent detection module generates product feature images based on the feature vectors output by the AD sampling module, and processes the product feature images to distinguish the differences from the standard qualified product feature images, wherein the processing includes image preprocessing, image feature extraction and target recognition detection; the judgment module judges the output results of the intelligent detection module, outputs the detection results, and then notifies the execution mechanism to make corresponding response actions.

10. The system of claim 9, wherein, The system further comprises a threshold detection module parallel to the intelligent detection module, which uses a preset threshold as the judgment standard, and if the sampling signal exceeds the threshold, it is determined that foreign matter is detected.

Images