A precious metal color recognition and detection system based on machine vision

Abstract

The invention relates to the technical field of image recognition and detection, and discloses a precious metal color recognition and detection system based on machine vision. The environmental parameters during precious metal detection are collected in real time by an environmental data collection module; the surface image of the precious metal is collected in real time by an image collection module; the surface image is preprocessed by an image preprocessing module, and color features, texture features and reflection features are extracted; a theoretical color value is obtained according to the color features, texture features and emission features by a component analysis module, the obtained theoretical color value is compared with a preset color value interval, and the theoretical color value is judged to be normal, to be confirmed or abnormal according to the comparison result; the theoretical color value to be confirmed is further analyzed by an inspection and analysis module, and an environmental abnormality value is obtained according to the environmental parameters; the obtained environmental abnormality value is compared with a preset environmental threshold value, and whether the theoretical color value is abnormal due to the environmental parameters is judged according to the comparison result.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition and detection, and in particular to a precious metal color recognition and detection system based on machine vision. Background Art

[0002] Precious metals, such as gold, silver, platinum, etc., are widely used in jewelry, investment, and industry due to their rarity and high value. Traditionally, the identification of the color and purity of precious metals relies on manual visual inspection or chemical analysis, which is both time-consuming and subjective. The time-consuming process will result in the inability to quickly obtain the color of the entire precious metal before smelting, resulting in a longer time before smelting, affecting the overall smelting efficiency; manual visual inspection will be subjective due to factors such as inconsistent vision, judgment standards, and the influence of lighting. Precious metals of the same color will get different results under the visual inspection of different inspectors, affecting the efficiency and accuracy of the color identification of precious metals.

[0003] Therefore, a system capable of automatically detecting the purity of precious metals is needed. Summary of the invention

[0004] The purpose of the present invention is to provide a precious metal color recognition and detection system based on machine vision to solve the above technical problems.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A precious metal color recognition and detection system based on machine vision, comprising:

[0007] Environmental data collection module, which collects environmental parameters during precious metal detection in real time through environmental monitoring equipment, including temperature, particulate matter content and light;

[0008] An image acquisition module, which acquires surface images of precious metals in real time through an image acquisition device;

[0009] The image preprocessing module preprocesses the surface image through a pre-trained image processing model and extracts color features, texture features, and reflection features;

[0010] The component analysis module inputs the real-time color features, texture features and emission features into the pre-trained color value prediction model, outputs the theoretical color value of the precious metal, compares the obtained theoretical color value with the preset color value range, and judges whether the theoretical color value is normal, to be confirmed or abnormal based on the comparison result;

[0011] The inspection and analysis module is used to further analyze the theoretical color value to be confirmed, obtain the environmental impact coefficient according to the environmental parameter processing, and fit the change curve of the environmental impact coefficient in unit time. And the reference change curve of the preset environmental impact coefficient For comparison, use the formula: Calculate the environmental anomaly value , where is the unit time length, , are the start time point and the end time point respectively. is the compensation coefficient;

[0012] The environmental outliers obtained Compare with the preset environmental threshold value, and judge whether the theoretical color value is abnormal due to environmental parameters based on the comparison result.

[0013] As a further technical solution, the specific process of judging whether the theoretical color value is abnormal is as follows:

[0014] The theoretical color value obtained Compared with the preset color value range [ ]Compare;

[0015] like ∈ , the theoretical color value is judged to be abnormal, and it is marked and kicked out of the subsequent precious metal processing sequence;

[0016] like ∈ , then the theoretical color value is judged to be normal and included in the subsequent precious metal processing sequence;

[0017] like ∈[ ], the theoretical color value needs to be confirmed and judged according to the analysis results of environmental parameters.

[0018] As a further technical solution, the process of obtaining the environmental impact coefficient according to the environmental parameter processing is:

[0019] By formula: ;

[0020] Calculate the environmental impact coefficient of the jth precious metal in the i-th unit time ;

[0021] In the formula, is the mean value of the environmental impact coefficient, The real-time particulate matter content in the current environment. is the predicted particulate matter content based on historical data, is the real-time illumination coefficient of the current environment, where , , B, C, and D are the light intensity, light uniformity, color temperature index, and glare index of the current environment, respectively. , , , is the preset scale factor, is the dimensionless coefficient, is the current environment real-time temperature, is the reference coefficient, which is selected and determined based on historical data and experimental data. is the temperature state value conversion function.

[0022] As a further technical solution, the obtained environmental anomaly values With preset environmental threshold The comparison process is:

[0023] like ≥ , it is judged that the current environment is abnormal;

[0024] like < , then the current environment is judged to be normal.

[0025] As a further technical solution, the process of judging whether the theoretical color value is abnormal due to environmental parameters according to the comparison result is as follows:

[0026] By formula: ;

[0027] Calculate the judgment value ;

[0028] In satisfying ∈[ ]and ≥ Under the conditions;

[0029] like > 0, it is judged that the environmental parameters cause the theoretical color value to be abnormal;

[0030] Otherwise, it is judged that non-environmental parameters cause the theoretical color value to be abnormal;

[0031] in, 0 is the judgment threshold, is a fixed conversion function, is the total number of unit time, For the Environmental anomaly value per unit time, is the mean color value.

[0032] As a further technical solution, the system further includes an early warning module, which is used to issue an early warning signal for precious metals with abnormal color.

[0033] As a further technical solution, the working process of the early warning module is:

[0034] when ∈ When the theoretical color value is abnormal, the current precious metal is marked and a secondary warning signal is issued;

[0035] when ∈[ ], a first-level warning signal needs to be issued;

[0036] like , it is determined that the current environment is abnormal, and the external fan is started to blow air to clean the current precious metal surface and image acquisition equipment;

[0037] If after blowing and cleaning, > 0, a level 1 yellow warning signal is issued;

[0038] Otherwise, a level one red warning signal will be issued.

[0039] As a further technical solution, the level of the second-level warning signal is higher than that of the first-level warning signal.

[0040] Beneficial effects of the present invention:

[0041] First, the environmental parameters and surface images around the precious metals waiting for color detection are collected in real time through the environmental data acquisition module and the image acquisition module respectively, and then the image processing model is pre-trained through historical data and machine learning algorithms, so as to extract color features, texture features and reflection features after preprocessing the surface image. According to the color features, texture features and reflection features, the theoretical color value of the current precious metal can be output through the color value prediction model, and the obtained theoretical color value is compared with the preset color value range. Since the structure of the theoretical color value is also affected by the environmental parameters, the theoretical color value is judged as normal, to be confirmed or abnormal according to the comparison result. If it is abnormal, it means that the theoretical color value is normal. If the color value is obviously abnormal and needs to be confirmed, it means that although the theoretical color value has certain abnormalities, it cannot be ruled out that the environmental parameters may cause the machine vision to misjudge. Therefore, the theoretical color value to be confirmed is further analyzed through the inspection and analysis module. The environmental impact coefficient is obtained according to the environmental parameter processing to obtain the environmental abnormality value. The obtained environmental abnormality value is compared with the preset environmental threshold value. According to the comparison result, it is judged whether the theoretical color value is abnormal due to the environmental parameters. In this way, it can be determined whether the abnormality of the theoretical color value is due to the poor color of the precious metal itself or the machine vision misjudgment due to abnormal environmental parameters, thereby improving the efficiency and accuracy of the color identification of precious metals. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present invention will be further described below in conjunction with the accompanying drawings.

[0043] Figure 1 It is a structural block diagram of the present invention. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0045] See also Figure 1 As shown, the present invention is a precious metal color recognition and detection system based on machine vision, comprising:

[0046] Environmental data collection module, which collects environmental parameters during precious metal detection in real time through environmental monitoring equipment, including temperature, particulate matter content and light;

[0047] An image acquisition module, which acquires surface images of precious metals in real time through an image acquisition device;

[0048] The image preprocessing module preprocesses the surface image through a pre-trained image processing model and extracts color features, texture features, and reflection features;

[0049] The component analysis module inputs the real-time color features, texture features and emission features into the pre-trained color value prediction model, outputs the theoretical color value of the precious metal, compares the obtained theoretical color value with the preset color value range, and judges whether the theoretical color value is normal, to be confirmed or abnormal based on the comparison result;

[0050] The inspection and analysis module is used to further analyze the theoretical color value to be confirmed, obtain the environmental impact coefficient according to the environmental parameter processing, and fit the change curve of the environmental impact coefficient in unit time. And the reference change curve of the preset environmental impact coefficient For comparison, use the formula: Calculate the environmental anomaly value , where is the unit time length, , are the start time point and the end time point respectively. is the compensation coefficient;

[0051] The environmental outliers obtained Compare with the preset environmental threshold value, and judge whether the theoretical color value is abnormal due to environmental parameters based on the comparison result.

[0052] The environmental outliers obtained With preset environmental threshold The comparison process is:

[0053] like ≥ , it is judged that the current environment is abnormal;

[0054] like < , then the current environment is judged to be normal.

[0055] The process of judging whether the theoretical color value is abnormal due to environmental parameters based on the comparison results is as follows:

[0056] By formula: ;

[0057] Calculate the judgment value ;

[0058] In satisfying ∈[ ]and ≥ Under the conditions;

[0059] like > 0, it is judged that the environmental parameters cause the theoretical color value to be abnormal;

[0060] Otherwise, it is judged that non-environmental parameters cause the theoretical color value to be abnormal;

[0061] in, 0 is the judgment threshold, is a fixed conversion function, is the total number of unit time, For the Environmental anomaly value per unit time, is the mean color value.

[0062] In this embodiment, the environmental parameters and surface images around the precious metal to be detected are first collected in real time by the environmental data collection module and the image collection module, and then the image processing model is pre-trained by the historical data and the machine learning algorithm, so as to extract the color features, texture features and reflection features after pre-processing the surface image. According to the color features, texture features and reflection features, the theoretical color value of the current precious metal can be output through the color value prediction model, and the obtained theoretical color value is compared with the preset color value range. Since the structure of the theoretical color value is also affected by the environmental parameters, the theoretical color value is judged to be normal, to be confirmed or abnormal according to the comparison result, and an abnormality is indicated. If the theoretical color value is obviously abnormal and needs to be confirmed, it means that although there is a certain abnormality in the theoretical color value, the possibility of misjudgment of machine vision due to environmental parameters cannot be ruled out. Therefore, the theoretical color value to be confirmed is further analyzed through the inspection and analysis module, and the environmental impact coefficient is obtained according to the environmental parameter processing to obtain the environmental abnormality value. The obtained environmental abnormality value is compared with the preset environmental threshold value, and the comparison result is used to determine whether the theoretical color value is abnormal due to the environmental parameters; thereby determining whether the abnormality in the theoretical color value is due to the poor color of the precious metal itself or the machine vision misjudgment due to abnormal environmental parameters, thereby improving the efficiency and accuracy of the color identification of precious metals.

[0063] The specific method for judging the abnormality of environmental parameters is to use the formula Calculate the environmental anomaly value Obviously, combining the extreme value of the environmental impact coefficient within a unit time and The cumulative difference can more accurately reflect the abnormal state of the environment and play a role in accurately judging the abnormal environment. is the compensation coefficient, which is formulated based on the experimental test data and will not be described in detail; then the obtained environmental abnormal value is With preset environmental threshold Compare, if ≥ , then it is judged that the current environment is abnormal; if < , then the current environment is normal. Obviously, when the environment is abnormal, it may be that there is floating dust in the camera of precious metals or image acquisition equipment, which leads to recognition errors. Therefore, after blowing dust with an external fan, the formula is used again. ; Calculate and obtain the judgment value ; It is obvious that The larger it is, the greater the fluctuation of environmental abnormal values, which means the greater the environmental problem. The larger the value, the greater the deviation between the theoretical color value and the average color value, so the greater the influence of environmental parameters on the color value; at the same time, if the ∈[ ]and ≥ Under the condition of > 0, it is judged that the environmental parameters cause the abnormal theoretical color value; otherwise, it is judged that the non-environmental parameters cause the abnormal theoretical color value; after eliminating the dust factor, it can be inferred that there are problems with temperature and light, which lead to the abnormal theoretical color value, providing a clear direction for the subsequent adjustment of environmental parameters.

[0064] The specific process of judging whether the theoretical color value is abnormal is as follows:

[0065] The theoretical color value obtained Compared with the preset color value range [ ]Compare;

[0066] like ∈ , the theoretical color value is judged to be abnormal, and it is marked and kicked out of the subsequent precious metal processing sequence;

[0067] like ∈ , then the theoretical color value is judged to be normal and included in the subsequent precious metal processing sequence;

[0068] like ∈[ ], the theoretical color value needs to be confirmed and judged according to the analysis results of environmental parameters.

[0069] In this embodiment, a specific method for judging whether the theoretical color value is abnormal is provided. ∈ , the theoretical color value is judged to be abnormal, marked and kicked out of the subsequent precious metal processing sequence; ∈ , then the theoretical color value is judged to be normal and included in the subsequent precious metal processing sequence; when ∈[ ], the theoretical color value needs to be confirmed and judged according to the analysis results of environmental parameters.

[0070] The process of obtaining the environmental impact coefficient according to environmental parameter processing is:

[0071] By formula: ;

[0072] Calculate the environmental impact coefficient of the jth precious metal in the i-th unit time ;

[0073] In the formula, is the mean value of the environmental impact coefficient, The real-time particulate matter content in the current environment. is the predicted particulate matter content based on historical data, is the real-time illumination coefficient of the current environment, where , , B, C, and D are the light intensity, light uniformity, color temperature index, and glare index of the current environment, respectively. , , , is the preset scale factor, is the dimensionless coefficient, is the current environment real-time temperature, is the reference coefficient, which is selected and determined based on historical data and experimental data. is the temperature state value conversion function, which is obtained by fitting the experimental data.

[0074] In this embodiment, a method for obtaining an environmental impact coefficient is provided, specifically, by the formula: Calculate the environmental impact coefficient of the jth precious metal in the i-th unit time Obviously, if The larger the value, the greater the difference between the current real-time particulate matter content and the particulate matter content predicted based on historical data. Therefore, the greater the air quality, the more likely it is that impurities on the surface of the precious metal to be tested will cause color detection anomalies. It is calculated based on the overall changes in temperature and overall changes in light. , combined with the mean value of the environmental impact coefficient , which can comprehensively reflect the change data of the environmental impact coefficient and provide accurate data support for judging whether the abnormal theoretical color value is due to environmental problems.

[0075] The system also includes an early warning module, which is used to issue an early warning signal to precious metals with abnormal color.

[0076] The working process of the early warning module is:

[0077] when ∈ When the theoretical color value is abnormal, the current precious metal is marked and a secondary warning signal is issued;

[0078] When O∈ A first-level warning signal needs to be issued;

[0079] like ≥ , it is determined that the current environment is abnormal, and the external fan is started to blow air to clean the current precious metal surface and image acquisition equipment;

[0080] If after blowing and cleaning, > 0, a level 1 yellow warning signal is issued;

[0081] Otherwise, a level one red warning signal will be issued.

[0082] The level of the second-level warning signal is higher than that of the first-level warning signal.

[0083] In this embodiment, a method for timely warning after an abnormality occurs in the precious metal color detection is provided. Specifically, when ∈ When the theoretical color value is abnormal, the current precious metal is marked and a secondary warning signal is issued; ∈[ ], a first-level warning signal needs to be issued; if ≥ , it is judged that the current environment is abnormal, and the external fan is started to blow air to clean the current precious metal surface and image acquisition equipment; if after blowing and cleaning, > 0, a first-level yellow warning signal is issued; otherwise, a first-level red warning signal is issued. The first-level red warning signal and the first-level yellow warning signal can directly feedback whether it is more likely that the precious metal itself is abnormal or the environment is abnormal, which leads to abnormal theoretical color value, indicating the direction for subsequent investigation and detection.

[0084] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A precious metal color recognition and detection system based on machine vision, characterized in that: include: Environmental data acquisition module, which collects environmental parameters during precious metal detection in real time through environmental monitoring equipment, including temperature, particulate matter content and light; An image acquisition module, which acquires surface images of precious metals in real time through an image acquisition device; The image preprocessing module preprocesses the surface image through a pre-trained image processing model and extracts color features, texture features, and reflection features; The component analysis module inputs the real-time color features, texture features and emission features into the pre-trained color value prediction model, outputs the theoretical color value of the precious metal, compares the obtained theoretical color value with the preset color value range, and judges whether the theoretical color value is normal, to be confirmed or abnormal based on the comparison result; The inspection and analysis module is used to further analyze the theoretical color value to be confirmed, obtain the environmental impact coefficient according to the environmental parameter processing, and fit the change curve of the environmental impact coefficient in unit time. And the reference change curve of the preset environmental impact coefficient For comparison, use the formula: Calculate the environmental anomaly value , where is the unit time length, , are the start time point and the end time point respectively. is the compensation coefficient; The environmental outliers obtained Compare with the preset environmental threshold, and judge whether the theoretical color value is abnormal due to environmental parameters according to the comparison result; The specific process of judging whether the theoretical color value is abnormal is as follows: The theoretical color value obtained Compared with the preset color value range [ ]Compare; like ∈ , the theoretical color value is judged to be abnormal, and it is marked and kicked out of the subsequent precious metal processing sequence; like ∈ , then the theoretical color value is judged to be normal and included in the subsequent precious metal processing sequence; like ∈[ ], the theoretical color value needs to be confirmed and judged according to the analysis results of environmental parameters; The process of obtaining the environmental impact coefficient according to environmental parameter processing is: By formula: ; Calculate the environmental impact coefficient of the jth precious metal in the i-th unit time ; In the formula, is the mean value of the environmental impact coefficient, The real-time particulate matter content in the current environment. is the predicted particulate matter content based on historical data, is the real-time illumination coefficient of the current environment, where , , B, C, and D are the light intensity, light uniformity, color temperature index, and glare index of the current environment, respectively. , , , is the preset scale factor, is the dimensionless coefficient, is the current environment real-time temperature, is the reference coefficient, which is selected and determined based on historical data and experimental data. is the temperature state value conversion function; The process of judging whether the theoretical color value is abnormal due to environmental parameters based on the comparison results is as follows: By formula: ; Calculate the judgment value ; In satisfying ∈[ ]and ≥ Under the conditions; like > 0, it is judged that the environmental parameters cause the theoretical color value to be abnormal; Otherwise, it is judged that non-environmental parameters cause the theoretical color value to be abnormal; in, 0 is the judgment threshold, is a fixed conversion function, is the total number of unit time, For the Environmental anomaly value per unit time, is the mean color value.

2. The precious metal color identification and detection system based on machine vision according to claim 1 is characterized in that: The environmental outliers obtained With preset environmental threshold The comparison process is: like ≥ , it is judged that the current environment is abnormal; like < , then the current environment is judged to be normal.

3. The precious metal color identification and detection system based on machine vision according to claim 1 is characterized in that: The system also includes an early warning module, which is used to issue an early warning signal to precious metals with abnormal color.

4. The precious metal color identification and detection system based on machine vision according to claim 3 is characterized in that: The working process of the early warning module is: when ∈ When the theoretical color value is abnormal, the current precious metal is marked and a secondary warning signal is issued; when ∈[ ], a first-level warning signal needs to be issued; like ≥ , it is judged that the current environment is abnormal, and the external fan is started to blow air to clean the current precious metal surface and image acquisition equipment; If after blowing and cleaning, > 0, a level 1 yellow warning signal is issued; Otherwise, a level one red warning signal will be issued.

5. The precious metal color recognition and detection system based on machine vision according to claim 4 is characterized in that: The level of the second-level warning signal is higher than that of the first-level warning signal.

Images