A light-sensitive currency counting machine with image recognition function

Abstract

The present invention discloses a light-sensing banknote detector with an image recognition function, which relates to the technical field of banknote detectors and solves the problem that when the banknote is aged or the fluorescent area is blurred, the detection accuracy has a large error. The present invention confirms and evaluates the light beam states at different points in the area to be detected of the detection object, locks the standard-reaching state of the corresponding points, and then performs relevant classification of the corresponding area straight lines based on the evaluation results of the corresponding points, confirms the numerical proportions associated with the corresponding straight lines, confirms the relevant partitions with relatively consistent features based on the top-down evaluation processing method of the numerical proportions, and performs detailed confirmation of the verification wavelengths for the confirmed relevant partitions, so that partitions with different features can match different verification wavelengths for associated verification, which can ensure the detection accuracy of the detection object in the later stage and improve the overall detection effect.

Description

Technical Field

[0001] The invention relates to the technical field of currency detectors, in particular to a light-sensing point currency detector with an image recognition function. Background Art

[0002] Banknotes use special fluorescent ink, which will emit a specific color of fluorescence under ultraviolet light; the banknote detector uses ultraviolet light to illuminate the banknote and detects the fluorescent reaction of the fluorescent ink on the banknote to determine the authenticity of the banknote. The fluorescent pattern of the real banknote is clear and the color is pure, while the fluorescence of the counterfeit banknote may be abnormal in color, uneven in brightness or blurred in pattern.

[0003] The application with publication number CN106023411A discloses a light-sensing banknote detector with image recognition function. The light-sensing banknote detector with image recognition function emits infrared laser of a certain wavelength through an infrared laser transmitter and irradiates it onto the fluorescent characters on the banknotes, which will cause the fluorescent characters to generate laser of a certain wavelength. The laser sensor can receive the laser generated by the fluorescent characters, and the laser is received and detected by a photoelectric converter. The photoelectric converter transmits the detection result to a controller, so that the authenticity of the banknotes can be quickly identified; the captured image is transmitted to an image comparison and recognition module through a camera, and the banknotes are verified through the image comparison and recognition module, and the banknote verification result is transmitted to the controller, so that the authenticity of the banknotes can be quickly identified; the two detection methods of laser detection and image recognition are integrated on the same device, the detection result is highly accurate, flexible and convenient to use, and the detection result has double insurance.

[0004] When the banknote detector is in use, it can quickly identify the authenticity of the corresponding banknote based on the corresponding fluorescence detection. However, in actual use, the banknote may be aged or the fluorescent area may be blurred, which will lead to large differences in detection accuracy when using the original specific wavelength of ultraviolet light for detection. Its detection accuracy needs to be improved. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a light-sensing banknote detector with an image recognition function, which solves the problem of large errors in detection accuracy when the banknote is aged or the fluorescent area is blurred.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a light-sensitive banknote detector with image recognition function, comprising:

[0007] The beam verification and analysis end uses a beam of a specific wavelength to perform a beam test on the detection object, determines the return beam associated with different points in the detection area, and evaluates the compliance status of the corresponding points based on the determined characteristic wavelength of the return beam. The specific method is as follows:

[0008] Based on the calibrated detection area within the detection object, confirm the return beams associated with different points in the detection area, and then determine the characteristic wavelength associated with the corresponding return beam based on the return beam waveform associated with the specified point:

[0009] Confirm the peak point of the return beam waveform, identify the horizontal and vertical distances between adjacent peak points, and mark the identified horizontal and vertical distances as L as the characteristic wavelength of the return beam waveform, and record the characteristic wavelength L associated with different points as L i , where i represents different points in the area to be detected;

[0010] Identify the characteristic wavelength L of the corresponding point i Satisfied: L i ≤Y1, where Y1 is the preset value. If it is satisfied, the corresponding point will be marked as a qualified point. Otherwise, the corresponding point will be marked as a non-qualified point.

[0011] The different points associated with the detection area are calibrated in turn as points that meet the standards and points that do not meet the standards;

[0012] The signal point calibration end confirms the proportion of the straight area for the different points calibrated in the detection area, and generates a numerical proportion sequence for the detection area based on the specific values ​​detected in real time from top to bottom, and transmits the confirmed numerical proportion sequence to the straight area partition end in the following manner:

[0013] Based on the marked area to be detected, confirm the points associated with the area to be detected from top to bottom, mark the points on the same horizontal line as the same-direction horizontal points, record this horizontal line as the same-direction horizontal line, and confirm the numerical value ratio associated with the same-direction horizontal line: confirm the total number of points G1 in the same-direction horizontal points, and then confirm the total number of substandard points G2 in the same-direction horizontal points, and use G2÷G1=ZB to confirm the numerical value ratio ZB associated with the same-direction horizontal line;

[0014] Then, according to the numerical proportions ZB associated with different horizontal lines in the same direction, the numerical proportion sequence of the area to be detected is confirmed according to the top-down sorting method, and the confirmed numerical proportion sequence is transmitted to the linear area partition end;

[0015] At the linear area partition end, based on the confirmed numerical proportion sequence, the difference of several groups of numerical proportions associated is confirmed, and the numerical proportions with lower differences are marked as similar proportions. Based on the relevant areas associated with the similar proportions, the similar average partition is confirmed from the area to be detected. The specific method is as follows:

[0016] Starting from the first numerical proportion of the numerical proportion sequence, identify the ratio difference between the first group of numerical proportions and the second group of numerical proportions, using: CY=|first group of numerical proportions-second group of numerical proportions|, confirm the ratio difference CY, and assess whether CY satisfies: CY≤Y2, where Y2 is a preset value;

[0017] If satisfied, the first group of numerical proportions and the second group of numerical proportions are classified as the same type of proportions, and the mean of the first group of numerical proportions and the second group of numerical proportions are determined, and the mean is compared with the third group of numerical proportions for the ratio difference confirmation. If the ratio difference is ≤Y2, the third group of numerical proportions are re-recorded into the same type of proportions. If the ratio difference is >Y2, the third group of numerical proportions are started from the third group of numerical proportions, and the related confirmation of the same type of proportions is performed afterwards;

[0018] If not, start from the second group of numerical proportions, identify the ratio difference between the second group of numerical proportions and the third group of numerical proportions, and based on the evaluation results of the ratio difference, confirm whether they belong to the same category of proportions;

[0019] Similarly, several groups of similar proportions in the numerical proportion sequence are confirmed in turn, and then the same-direction horizontal lines associated with each similar proportion are recorded as similar horizontal lines, and the areas associated with multiple groups of similar horizontal lines are recorded as similar average partitions, and the similar average partitions associated with each similar proportion are confirmed;

[0020] The calibration wavelength determination end determines the calibration wavelength associated with the corresponding similar average quantity partition based on the multiple groups of similar average quantity partitions confirmed in the area to be detected and the proportion of qualified points associated with each similar average quantity partition, and performs calibration through the execution end. The specific method is as follows:

[0021] Based on the confirmed similar average partition, the proportion of multiple groups of values ​​associated with this partition is analyzed by mean, the characteristic mean is confirmed, and the characteristic mean associated with different similar average partitions is calibrated as J k , where k represents different homogeneous mean partitions;

[0022] Based on the feature mean J associated with the corresponding partition k Confirm the UV wavelength B for the corresponding partition to be calibrated k , where B k =J k ×C1, where C1 is a fixed coefficient factor;

[0023] Then the UV wavelength B associated with different equal amount partitions k Perform synchronous verification, lock the verification partition, and associate the ultraviolet wavelength B of different equal-value partitions k Sorting is performed from top to bottom to confirm the wavelength sorting set;

[0024] Starting from the first ultraviolet wavelength in the wavelength sorting set, a numerical ascending segment and a numerical descending segment are determined, wherein the numerical ascending segment includes a partial numerical segment where the ultraviolet wavelength continues to increase and a partial numerical segment where the ultraviolet wavelength is equal, and the numerical descending segment is a partial numerical segment where the ultraviolet wavelength continues to decrease;

[0025] Confirm the multiple groups of similar average partitions associated with the rising segment of the value, and confirm the total area of ​​the multiple groups of similar average partitions. If the total area is ≥Y3, where Y3 is a preset value, then the multiple groups of similar average partitions are marked as execution partitions. If the total area is <Y3, then confirm the multiple groups of similar average partitions associated with the falling segment of the value, and confirm the total area of ​​the multiple groups of similar average partitions. If the total area is ≥Y3, then the multiple groups of similar average partitions associated with the falling segment of the value are marked as execution partitions. Otherwise, the determined multiple groups of similar average partitions are directly marked as execution partitions.

[0026] Preferably, the execution end verifies the same type of average quantity partition according to the determined execution partition and the ultraviolet light wavelength associated with different same type average quantity partitions within the execution partition, confirms the reflected fluorescent signal, and evaluates the authenticity of the detection object based on the presence or absence of the fluorescent signal.

[0027] The present invention provides a light-sensitive currency detector with image recognition function. Compared with the prior art, it has the following beneficial effects:

[0028] The present invention confirms and evaluates the light beam states at different points in the area to be detected of the detection object, locks the compliance state of the corresponding points, and then classifies the corresponding area straight lines based on the evaluation results of the corresponding points, confirms the numerical value ratio associated with the corresponding straight line, and confirms the relevant partitions with relatively consistent characteristics based on the top-down evaluation processing method of the numerical value ratio, and performs detailed confirmation of the verification wavelengths for the confirmed relevant partitions, so that partitions with different characteristics can match different verification wavelengths for correlation verification, which can ensure the detection accuracy of the detection object in the later stage and improve its overall detection effect;

[0029] In order to further ensure the accuracy during the detection process, a numerical evaluation and analysis is performed on the confirmed calibration wavelength to confirm the numerical rising segment or numerical falling segment. Based on the relevant proportion of the corresponding numerical rising segment or numerical falling segment, a reasonable selection is made. This ensures that the corresponding detection object does not need to change the calibration wavelength back and forth during the detection process. A single group change method is maintained, thereby further improving the overall detection accuracy of the detection object. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the principle framework of the present invention;

[0031] Figure 2 It is a schematic diagram for determining the target point of the present invention. DETAILED DESCRIPTION

[0032] 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.

[0033] First embodiment

[0034] See also Figure 1 , the present application provides a light-sensitive banknote detector with image recognition function, comprising a light beam verification and analysis end, a signal point position calibration end, a straight area partition end, a verification wavelength determination end and an execution end, wherein the light beam verification and analysis end, the signal point position calibration end, the straight area partition end, the verification wavelength determination end and the execution end are electrically connected from the output node to the input node in sequence;

[0035] Among them, at the beam verification and analysis end, a beam of a specific wavelength is used to perform a beam test on the detection object, and the return beam associated with different points in the area to be detected is determined, and based on the characteristic wavelength of the determined return beam, the compliance status of the corresponding points is evaluated. Specifically, the specific wavelength beam used here is 530nm green light, and its green light can effectively detect whether the pattern is blurred, and can effectively confirm the edge clarity and line integrity of the detection object. When the detection object is placed on the specified platform, the operating system can automatically lock the detection area, and the detection area is the fluorescent area associated with the corresponding banknote. The compliance status of the corresponding point can be directly determined based on the point numerical expression status of the fluorescent area. This method can effectively test aging banknotes or deformed and blurred banknotes. There is a window for placing the corresponding banknote in the banknote counting machine. Through the internal roller shaft, the banknote can automatically rotate to the detection area. Based on the detection of the corresponding beam, the fluorescent area associated with the banknote is effectively tested to evaluate the authenticity of the corresponding aging banknote.

[0036] Among them, the specific method of assessing the compliance status of the corresponding points is:

[0037] Based on the calibrated area to be detected (that is, the fluorescent area) in the detection object, confirm the return beam associated with different points in the area to be detected (that is, the specific beam reflected back), and then based on the return beam waveform associated with the specified point, determine the characteristic wavelength associated with the corresponding return beam:

[0038] Confirm the peak point of the return beam waveform (that is, the waveform point with the strongest parameter), identify the horizontal and vertical distances between adjacent peak points, and mark the identified horizontal and vertical distances as L as the characteristic wavelength of the return beam waveform, and record the characteristic wavelength L associated with different points as L i , where i represents different points in the area to be detected;

[0039] Identify the characteristic wavelength L of the corresponding point i Satisfied: L i ≤Y1, where Y1 is a preset value, and its specific value is determined by the operator based on experience. If it is satisfied, the corresponding point will be calibrated as a qualified point, otherwise, the corresponding point will be calibrated as a non-qualified point (under the green light detection of a specific wavelength of 530nm, the shorter the wavelength of the light beam reflected by the corresponding point, the stronger it is, and thus the more obvious the fluorescence associated with the point in the corresponding area is, otherwise, the more blurred the fluorescence associated with the corresponding point is);

[0040] For different points associated with the detection area, specific calibration is carried out in turn for the points that meet the standards and the points that do not meet the standards.

[0041] Among them, the signal point calibration end confirms the proportion of the straight area of ​​different points calibrated in the detection area, and generates a numerical proportion sequence for the detection area based on the specific values ​​detected in real time from top to bottom, and transmits the confirmed numerical proportion sequence to the straight area partition end, wherein the numerical proportion sequence is generated in the following manner:

[0042] Based on the marked area to be detected, confirm the points associated with the area to be detected from top to bottom, mark the points on the same horizontal line as the same-direction horizontal points, record this horizontal line as the same-direction horizontal line, and confirm the numerical value ratio associated with the same-direction horizontal line: confirm the total number of points G1 in the same-direction horizontal points, and then confirm the total number of substandard points G2 in the same-direction horizontal points, and use G2÷G1=ZB to confirm the numerical value ratio ZB associated with the same-direction horizontal line;

[0043] Then, according to the numerical proportions ZB associated with different horizontal lines in the same direction, the numerical proportion sequence of the area to be detected is confirmed according to the top-down sorting method, and the confirmed numerical proportion sequence is transmitted to the linear area partition end;

[0044] Specifically, the area to be detected is the corresponding fluorescent area, which belongs to a regular long strip area. Therefore, this area can confirm the horizontal related horizontal lines. Such related horizontal lines are sorted from top to bottom. During the sorting process, there are specific proportions of the points that meet the standards and the points that do not meet the standards on the corresponding horizontal lines. Based on the specific proportion of the points that do not meet the standards in the corresponding horizontal lines, the numerical proportions associated with different micro-areas in the corresponding area to be detected can be locked, and according to the specific numerical performance of different numerical proportions, the different partitions in the corresponding area to be detected can be locked.

[0045] Among them, at the linear area partition end, based on the confirmed numerical proportion sequence, the associated numerical proportions of several groups are confirmed to be different, the numerical proportions with lower differences are marked as similar proportions, and based on the relevant areas associated with the similar proportions, the similar average partitions are confirmed from the area to be detected. The specific method of confirmation is:

[0046] Starting from the first value proportion of the value proportion sequence, identify the difference in the ratio between the first group of value proportions and the second group of value proportions, using: CY=|first group of value proportions - second group of value proportions|, confirm the ratio difference CY, and assess whether CY satisfies: CY≤Y2, where Y2 is a preset value, and its specific value is determined by the operator based on experience;

[0047] If satisfied, the first group of numerical proportions and the second group of numerical proportions are classified as the same type of proportions, and the mean of the first group of numerical proportions and the second group of numerical proportions are determined, and the mean is compared with the third group of numerical proportions for the ratio difference confirmation. If the ratio difference is ≤Y2, the third group of numerical proportions are re-recorded into the same type of proportions. If the ratio difference is >Y2, the third group of numerical proportions are started from the third group of numerical proportions, and the related confirmation of the same type of proportions is performed afterwards;

[0048] If not, start from the second group of numerical proportions, identify the ratio difference between the second group of numerical proportions and the third group of numerical proportions, and based on the evaluation results of the ratio difference, confirm whether they belong to the same category of proportions;

[0049] Similarly, several groups of similar proportions in the numerical proportion sequence are confirmed in turn, and then the same-direction horizontal line associated with each similar proportion is recorded as a similar horizontal line, and the area associated with multiple groups of similar horizontal lines is recorded as a similar average partition, and the similar average partition associated with each similar proportion is confirmed. Specifically, the confirmed numerical proportion sequence is {ZB1, ZB2, ZB3, ZB4, ..., ZBn}. If the ratio difference between ZB1 and ZB2 does not meet the evaluation criteria, it will not be recorded. Starting from ZB2, the ratio difference is confirmed afterwards. When ZB2 and ZB3 belong to the same proportion, the mean ratio between ZB2 and ZB3 is confirmed, and then the ratio difference of ZB4 is confirmed based on the confirmed mean ratio. If they still belong to the same proportion, the mean of ZB2, ZB3 and ZB4 is confirmed, and the ratio difference of the confirmed mean and the corresponding ZB5 is confirmed. Similarly, multiple groups of similar proportions associated in the numerical proportion sequence are confirmed in turn to ensure the confirmation effect and improve the confirmation accuracy.

[0050] Second embodiment

[0051] In the specific implementation process of this embodiment, compared with the above embodiment, this embodiment is mainly used in the execution stage, and the subsequent calibration wavelength determination end and the execution end perform related execution;

[0052] Among them, combined Figure 2 , the verification wavelength determination end, based on the multiple groups of similar average quantity partitions confirmed in the area to be detected, and based on the proportion of standard points associated with each similar average quantity partition, confirms the verification wavelength associated with the corresponding similar average quantity partition (ultraviolet light is used for verification here, and green light is used to evaluate the aging degree of the detection object in the first embodiment. This embodiment is based on ultraviolet light to verify the authenticity of the detection object), wherein the specific method of confirming the verification wavelength is:

[0053] Based on the confirmed similar average partition, the proportion of multiple groups of values ​​associated with this partition is analyzed by mean, the characteristic mean is confirmed, and the characteristic mean associated with different similar average partitions is calibrated as J k , where k represents different homogeneous mean partitions;

[0054] Based on the feature mean J associated with the corresponding partition k Confirm the UV wavelength B for the corresponding partition to be calibrated k , where B k =J k ×C1, where C1 is a fixed coefficient factor, which is prepared in advance by relevant operators based on experience. The more the overall proportion of qualified points is, the longer the associated ultraviolet wavelength is; the fewer the overall proportion of qualified points is, the shorter the associated ultraviolet wavelength is. By targeting each one, the overall fluorescence calibration effect of the corresponding area can be effectively guaranteed;

[0055] Then the UV wavelength B associated with different equal amount partitions k Perform synchronous verification, lock the verification partition, and associate the ultraviolet wavelength B of different equal-value partitions k Sorting is performed from top to bottom to confirm the wavelength sorting set;

[0056] Starting from the first ultraviolet wavelength in the wavelength sorting set, a numerical ascending segment and a numerical descending segment are determined, wherein the numerical ascending segment includes a partial numerical segment where the ultraviolet wavelength continues to increase and a partial numerical segment where the ultraviolet wavelength is equal, and the numerical descending segment is a partial numerical segment where the ultraviolet wavelength continues to decrease;

[0057] Confirm the multiple groups of similar average partitions associated with the rising segment of the value, and confirm the total area of ​​the multiple groups of similar average partitions. If the total area is ≥Y3, where Y3 is a preset value, and its specific value is determined by the operator based on experience, then the multiple groups of similar average partitions are marked as execution partitions. If the total area is <Y3, then confirm the multiple groups of similar average partitions associated with the falling segment of the value, and confirm the total area of ​​the multiple groups of similar average partitions. If the total area is ≥Y3, then the multiple groups of similar average partitions associated with the falling segment of the value are marked as execution partitions. Otherwise, the determined multiple groups of similar average partitions are directly marked as execution partitions.

[0058] Among them, the execution end verifies the same type of average quantity partition according to the determined execution partition and the ultraviolet light wavelength associated with different same type average quantity partitions in the execution partition, confirms the reflected fluorescent signal, and evaluates the authenticity of the detection object based on the presence or absence of the fluorescent signal (if the fluorescent signal exists, it is true, and if the fluorescent signal does not exist, it is false).

[0059] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0060] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A light-sensitive banknote detector with image recognition function, characterized in that: include: The beam verification and analysis end uses a beam of a specific wavelength to perform a beam test on the detection object, determines the return beam associated with different points in the area to be detected, and evaluates the compliance status of the corresponding points based on the determined characteristic wavelength of the return beam; The signal point calibration end confirms the proportion of the straight area for different points calibrated in the area to be detected, and generates a numerical proportion sequence for the area to be detected based on the specific values ​​detected in real time from top to bottom, and transmits the confirmed numerical proportion sequence to the straight area partition end; At the linear area partition end, based on the confirmed numerical proportion sequence, the associated numerical proportions of several groups are confirmed to be different, the numerical proportions with lower differences are marked as similar proportions, and based on the relevant areas associated with the similar proportions, the similar average partitions are confirmed from the area to be detected; The calibration wavelength determination end determines the calibration wavelength associated with the corresponding similar average quantity partition based on the multiple groups of similar average quantity partitions confirmed in the area to be detected and the proportion of qualified points associated with each similar average quantity partition, and performs calibration through the execution end. The specific method is as follows: Based on the confirmed similar average partition, the proportion of multiple groups of values ​​associated with this partition is analyzed by mean, the characteristic mean is confirmed, and the characteristic mean associated with different similar average partitions is calibrated as J k , where k represents different homogeneous mean partitions; Based on the feature mean J associated with the corresponding partition k Confirm the UV wavelength B for the corresponding partition to be calibrated k , where B k =J k ×C1, where C1 is a fixed coefficient factor.

2. The optical banknote detector with image recognition function according to claim 1, characterized in that: The specific method of evaluating the compliance status of the corresponding points at the beam verification and analysis end is as follows: Based on the calibrated detection area within the detection object, confirm the return beams associated with different points in the detection area, and then determine the characteristic wavelength associated with the corresponding return beam based on the return beam waveform associated with the specified point: Confirm the peak point of the return beam waveform, identify the horizontal and vertical distances between adjacent peak points, and mark the identified horizontal and vertical distances as L as the characteristic wavelength of the return beam waveform, and record the characteristic wavelength L associated with different points as L i , where i represents different points in the area to be detected; Identify the characteristic wavelength L of the corresponding point i Satisfied: L i ≤Y1, where Y1 is the preset value. If it is satisfied, the corresponding point will be marked as a qualified point. Otherwise, the corresponding point will be marked as a non-qualified point. For different points associated with the detection area, specific calibration is carried out in turn for the points that meet the standards and the points that do not meet the standards.

3. The optical banknote detector with image recognition function according to claim 1, characterized in that: The signal point calibration end generates a numerical ratio sequence in the following manner: Based on the marked area to be detected, confirm the points associated with the area to be detected from top to bottom, mark the points on the same horizontal line as the same-direction horizontal points, record this horizontal line as the same-direction horizontal line, and confirm the numerical value ratio associated with the same-direction horizontal line: confirm the total number of points G1 in the same-direction horizontal points, and then confirm the total number of substandard points G2 in the same-direction horizontal points, and use G2÷G1=ZB to confirm the numerical value ratio ZB associated with the same-direction horizontal line; Then, according to the numerical proportions ZB associated with different horizontal lines in the same direction, the numerical proportion sequence of the area to be detected is confirmed according to the top-down sorting method, and the confirmed numerical proportion sequence is transmitted to the straight line area partition end.

4. The optical currency detector with image recognition function according to claim 1, characterized in that: The specific method of confirming the same type of average volume partition from the area to be detected is as follows: Starting from the first numerical proportion of the numerical proportion sequence, identify the ratio difference between the first group of numerical proportions and the second group of numerical proportions, using: CY=|first group of numerical proportions-second group of numerical proportions|, confirm the ratio difference CY, and assess whether CY satisfies: CY≤Y2, where Y2 is a preset value; If satisfied, the first group of numerical proportions and the second group of numerical proportions are classified as the same type of proportions, and the mean of the first group of numerical proportions and the second group of numerical proportions are determined, and the mean is compared with the third group of numerical proportions for the ratio difference confirmation. If the ratio difference is ≤Y2, the third group of numerical proportions are re-recorded into the same type of proportions. If the ratio difference is >Y2, the third group of numerical proportions are started from the third group of numerical proportions, and the related confirmation of the same type of proportions is performed afterwards; If not, start from the second group of numerical proportions, identify the ratio difference between the second group of numerical proportions and the third group of numerical proportions, and based on the evaluation results of the ratio difference, confirm whether they belong to the same category of proportions; Several groups of similar proportions in the numerical proportion sequence are confirmed in turn, and then the same-direction horizontal line associated with each similar proportion is recorded as a similar horizontal line, the area associated with multiple groups of similar horizontal lines is recorded as a similar average partition, and the similar average partition associated with each similar proportion is confirmed.

5. The optical currency detector with image recognition function according to claim 1, characterized in that: The calibration wavelength determining end further comprises: Then the UV wavelength B associated with different equal amount partitions k Perform synchronous verification, lock the verification partition, and associate the UV wavelength B of different equal-value partitions k Sorting is performed from top to bottom to confirm the wavelength sorting set; Starting from the first ultraviolet wavelength in the wavelength sorting set, a numerical ascending segment and a numerical descending segment are determined, wherein the numerical ascending segment includes a partial numerical segment where the ultraviolet wavelength continues to increase and a partial numerical segment where the ultraviolet wavelength is equal, and the numerical descending segment is a partial numerical segment where the ultraviolet wavelength continues to decrease; Confirm the multiple groups of similar average partitions associated with the rising segment of the value, and confirm the total area of ​​the multiple groups of similar average partitions. If the total area is ≥Y3, where Y3 is a preset value, then the multiple groups of similar average partitions are marked as execution partitions. If the total area is <Y3, then confirm the multiple groups of similar average partitions associated with the falling segment of the value, and confirm the total area of ​​the multiple groups of similar average partitions. If the total area is ≥Y3, then the multiple groups of similar average partitions associated with the falling segment of the value are marked as execution partitions. Otherwise, the determined multiple groups of similar average partitions are directly marked as execution partitions.

6. The optical currency detector with image recognition function according to claim 5, characterized in that: The execution end verifies the same type average quantity partition according to the determined execution partition and the ultraviolet light wavelength associated with different same type average quantity partitions in the execution partition, confirms the reflected fluorescent signal, and evaluates the authenticity of the detection object based on the presence or absence of the fluorescent signal.

Images