JP2018151338A - Article inspection device and inspection condition switch method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow inspection conditions appropriate for the category of inspection objects to be correctly set.SOLUTION: Of inspection conditions including operation conditions of transmitting/receiving an electromagnetic wave required for inspection, and processing conditions of generation and determination processing of image data, the inspection condition is registered, in inspection condition registration means 60, that includes a processing condition of the image data common in the operation condition of transmitting/receiving the electromagnetic wave and likely to change the common inspection condition depending upon any of a category, passage posture and thickness of an inspection object, and distinguishable feature data about any of the category, passage posture and thickness of the inspection object is configured to be registered in feature data registration means 90 on the basis of a reception output of the electromagnetic wave to be obtained when the inspection object passes through a passage path under the common operation condition. Inspection condition switch means 100 is configured to: when the inspection object of the feature data different from the last passing inspection object passes, read the inspection condition about the category, the passage posture, and the thickness to be distinguished by the feature data; and update a generation processing condition of the image data with respect to the object to be inspected.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a technique for accurately setting inspection conditions suitable for the type of inspection object in an article inspection apparatus for examining the presence of defects such as foreign matters mixed in products such as foods and missing contents. .

  In factories that produce food, etc., use an article inspection device to check for defects such as foreign matter contamination or missing items in the product, and automatically inspect the article while it is being conveyed by a conveyor, etc. To separate good and defective products.

  In particular, in products such as food, it is necessary to strictly inspect the presence of foreign matters such as metals and plastics. In order to cope with this, in recent years, article inspection apparatuses using X-rays have been realized.

  In general, an article inspection apparatus using X-rays emits X-rays having a width in a direction orthogonal to the passage direction of a passage through the inspection object, and inspects the X-rays transmitted through the inspection object. By receiving a plurality of X-ray sensors arranged in a direction orthogonal to the object passing direction, obtaining image information indicating a difference in transmittance of each part of the object to be inspected with respect to X-rays, and performing various processes on the image information It is determined whether or not foreign matter is mixed in, contents are missing, missing or the like.

  In such an article inspection apparatus using X-rays, various inspection conditions such as the intensity of X-rays used for inspection, the algorithm of image processing, and the threshold value used therefor vary depending on the type of inspection object.

  For this reason, the most appropriate inspection condition for each type of inspection object is registered in advance in the apparatus, and the operator selects an inspection object to be inspected from among the registered types, The apparatus is operated under the inspection conditions of the selected product type.

  For example, Patent Document 1 discloses an inspection apparatus that registers optimum inspection conditions for each product type and operates the device under the product inspection conditions selected by the operator.

JP2007-232586

  However, even inspected items of the same variety, the inspected items that are both in a standing posture and a sleeping state when passing through the passage, or contents such as frozen meat, for example There are inspection objects that cause large variations in the thickness in the X-ray transmission direction, and when inspection is performed under the same inspection conditions for such types of inspection objects, the difference in the passing posture and the thickness There are cases where correct test results cannot be obtained due to differences.

  In addition, for example, “Meat Gyoza” and “Vegetable Gyoza”, “Meat Man” and “Amman”, “Shumai” and “Gyoza”, etc. In the case of carrying in a mixture, if the inspection is performed under the same inspection condition, a correct inspection result may not be obtained due to the difference in material and moisture content.

  One way to solve the above problem is to register the optimum inspection conditions obtained in advance for different passage postures and thickness differences for the same product, and immediately before the inspection object is carried in A method of detecting the passing posture and thickness of the inspection object and using the inspection condition corresponding to the detection result is conceivable.

  However, this method requires separate detection means for detecting the passing posture of the inspection object before carrying in or detecting the thickness, and in addition, these detection means cannot identify the above-mentioned approximate products, It is not possible to carry in when mixed with similar varieties.

  The present invention solves this problem, and inspects the inspected object not only in the passing posture and thickness difference of the inspected object of the same type, but also in the case of bringing in similar types mixedly. It is an object of the present invention to provide an article inspection apparatus capable of setting conditions correctly and an inspection condition switching method thereof.

In order to achieve the above object, an article inspection apparatus according to claim 1 of the present invention comprises:
An electromagnetic wave transmission unit (22) for outputting electromagnetic waves of different wavelength regions in a passage path through which the inspection object passes;
An electromagnetic wave receiver (30) for receiving an electromagnetic wave output from the electromagnetic wave generator to the passage and transmitted through the object to be inspected;
Image data generating means (40) for generating image data of an object to be inspected with respect to electromagnetic waves in the plurality of different wavelength regions by signal processing on the output of the electromagnetic wave receiving unit;
A determination unit (50) for performing a determination process of the quality of the inspection object from the image data generated by the image data generation unit;
Among the inspection conditions including the operation condition of the electromagnetic wave transmission unit necessary for the inspection of the inspection object, the operation condition of the electromagnetic wave reception unit, the processing condition of the image data of the image data generation unit, and the determination processing condition of the determination unit The operation condition of the electromagnetic wave transmission unit and the electromagnetic wave reception unit is the same, and the image data generation unit may change depending on at least one of the type of inspection object, the passing posture, and the difference in thickness. Inspection conditions including image data processing conditions are pre-registered inspection condition registration means (60);
Based on the output signal of the electromagnetic wave reception unit obtained when the inspection object passes through the passage when the electromagnetic wave transmission unit and the electromagnetic wave reception unit are in the common operating condition, the type of the inspection object, A feature data acquisition means (80) for acquiring feature data capable of identifying either a passing posture or a difference in thickness;
Feature data registration means (90) in which feature data acquired in advance by the feature data acquisition means for the inspection object to be inspected is registered;
Each time the inspection object to be inspected passes through the passage, it receives feature data newly acquired by the feature data acquisition means, and the newly acquired feature data is registered in the feature data registration means and If the feature data acquired for the inspection object passed immediately before is different from the feature data acquired, the inspection condition registration means indicates the inspection condition for any of the product type, the passing posture, and the thickness identified by the newly acquired feature data. And inspection condition switching means (100) for updating the inspection condition including the processing condition of the image data of the image data generating means for the inspection object to be inspected with the read inspection condition.

Moreover, the article inspection apparatus according to claim 2 of the present invention is the article inspection apparatus according to claim 1,
A plurality of different wavelength regions of the electromagnetic wave output from the electromagnetic wave transmitting means are in an X-ray region.

Moreover, the article inspection apparatus according to claim 3 of the present invention is the article inspection apparatus according to claim 1 or 2,
The feature data acquired by the feature data acquisition means includes the shape of the density distribution or the total density of the image of the object to be inspected obtained when the electromagnetic wave transmitting unit and the electromagnetic wave receiving unit are in the common operating condition, or on the image. At least one of the areas is included.

Moreover, the article inspection apparatus according to claim 4 of the present invention is the article inspection apparatus according to any one of claims 1 to 3,
The characteristic data acquired by the characteristic data acquisition means includes the transmittance for electromagnetic waves in the plurality of wavelength regions of the object to be inspected obtained when the electromagnetic wave transmitting unit and the electromagnetic wave receiving unit are in the common operating condition. It is characterized by that.

Moreover, the inspection condition switching method of the article inspection apparatus according to claim 5 of the present invention is:
Output electromagnetic waves of different wavelength regions to the passages through which the inspection object passes, receive electromagnetic waves transmitted through the inspection object, and inspect the electromagnetic waves of different wavelength regions by signal processing on the received output An inspection condition switching method for an article inspection apparatus that generates image data of an object, and performs determination processing of the quality of the inspection object from the generated image data,
Among the inspection conditions including the electromagnetic wave transmission / reception operation conditions necessary for the inspection of the object to be inspected, the image data generation and the determination process processing conditions, the electromagnetic wave transmission / reception operation conditions are common and at least the target Registering the inspection conditions including the processing conditions of the image data that may be changed depending on the type of inspection object, the passing posture, or the difference in thickness;
Based on the received output obtained when the inspected object passes through the passage under the common operating conditions in advance, the type of the inspected object, the passing posture, or the difference in thickness is identified. Registering possible feature data,
Each time the inspection object to be inspected passes through the passage, the characteristic data is newly acquired, and the newly acquired characteristic data is registered and the characteristic data acquired for the inspection object that has passed immediately before If different, the inspection condition for any of the product type, the passing posture, and the thickness identified by the newly acquired feature data is read out, and the image data for the inspection object to be inspected under the read inspection condition And a step of updating the inspection condition including the above processing conditions.

  As described above, in the article inspection apparatus according to the present invention, among the inspection conditions including the operation conditions of electromagnetic wave transmission / reception necessary for the inspection of the inspection object and the processing conditions of image data generation and determination processing, the transmission / reception of electromagnetic waves is performed in advance. Register the inspection conditions including the processing conditions of the image data that can be changed depending on at least one of the kind of inspection object, the passing posture, and the difference in thickness. Based on the received output of the electromagnetic wave obtained when the inspection object passes through the passage under the common operating conditions of transmission and reception, identification of any difference in the type, passage posture, or thickness of the inspection object is possible. Possible feature data is registered, and each time the inspection object to be inspected passes through the passage, the newly acquired characteristic data has been registered and the inspection has passed immediately before the inspection object. Get about things Read out the inspection condition for any of the product type, the passing posture, and the thickness identified by the newly acquired feature data, and at least the inspection object to be inspected under the read inspection condition. Inspection conditions including image data generation processing conditions are updated.

  For this reason, the correct inspection conditions are set for the inspection object each time, even when there are differences in the posture and thickness of the inspection object of the same type, or when similar types are mixed in. And inspection of goods can be performed accurately.

Overall configuration diagram of an embodiment of the present invention The figure which shows the relationship between the pulse signal output from an X-ray sensor, and an area | region The block diagram of the principal part of embodiment of this invention The figure which shows the example of three types of image data obtained for every area | region of a crest value Diagram showing an example of registration of inspection conditions The figure which shows the example of the histogram of the density distribution of image data Diagram showing transmittance characteristics with respect to X-ray energy Diagram showing the relationship between passing posture and concentration distribution A diagram showing the relationship between the thickness of the contents of the inspection object and the concentration distribution

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 shows the overall configuration of an article inspection apparatus 20 to which the present invention is applied.

  The article inspection apparatus 20 includes a transport device 21, an electromagnetic wave transmission unit 22, an electromagnetic wave reception unit 30, an image data generation unit 40, a determination unit 50, an inspection condition registration unit 60, a product type switching unit 70, a feature data acquisition unit 80, and feature data. Registration means 90 and inspection condition switching means 100 are provided.

  The transport device 21 is for transporting the inspection object W in a predetermined direction (in the figure, the direction orthogonal to the paper surface). Generally, the inspection apparatus W is horizontally transported at a constant speed like a conveyor. However, it is not always necessary to use a conveyance device having a power source, and a method of sliding on a ramp using the weight of an object to be inspected or a method of dropping from above may be used.

  The electromagnetic wave transmission unit 22 outputs electromagnetic waves in a plurality of different wavelength regions to the passages through which the inspection object W passes. The electromagnetic wave used here may be an electromagnetic wave having an appropriate transmittance not only for the object to be inspected itself but also for its packaging material. For example, X-rays (including γ-rays from the shorter wavelength) are used. ), Infrared, microwave, etc. are candidates. Here, a case where the electromagnetic wave transmission unit 22 outputs X-rays will be described, but other electromagnetic waves can also be used.

  In this embodiment, the electromagnetic wave transmission unit 22 emits X-rays that spread in the width direction of the conveyance path from above the inspection object W conveyed by the conveyance device 21. Not limited to this, the light may be emitted from the side of the object W to the side.

  The electromagnetic wave transmission unit 22 includes a hot cathode X-ray tube that accelerates electrons emitted from a heated filament to collide with an anode target and emits X-rays as an X-ray source, and a lattice-controlled hot cathode X-ray. A tube is used and, in addition, a power source necessary for driving the X-ray tube is included.

  The energy of the X-ray photons output by the electromagnetic wave transmission unit 22 having the above structure is not constant and varies, and the energy of the X-ray photons depends on the wavelength of the X-rays. That is, the X-ray output from the electromagnetic wave transmission unit 22 includes a plurality of different wavelength regions. The X-ray energy output from the electromagnetic wave transmission unit 22 needs to be set in a range suitable for the inspection of the inspection object. This setting is generally performed by controlling the tube voltage and tube current applied to the X-ray tube. It is assumed that the parameters for each type necessary for these controls are registered in advance in an inspection condition registration unit 60 described later.

The electromagnetic wave receiving unit 30 includes a plurality of N X-ray sensors 31 _{1 to} 31 _N each having a function of receiving an X-ray and converting it into an electrical signal, and the plurality of N X-ray sensors 31 _{1 to} 31 _N are inspected. At the position where the X-rays transmitted through the object W are received, they are arranged in a line with almost no gap in the direction intersecting (orthogonal in this example) with the passing direction of the object W (the direction orthogonal to the paper surface).

As an actual apparatus, the plurality of N X-ray sensors 31 _{1 to} 31 _N have a single line sensor structure in which each of them is integrally connected, and is provided on the lower surface side of the conveyance path of the conveyance apparatus 21. Has been placed. Here, for example, assuming that the width of the X-ray sensors is 1 mm, and the gap between the X-ray sensors is negligibly small with respect to the width, and the width of the transport path for transporting the inspection object W is 200 mm, approximately 200 pieces. A line sensor having an X-ray sensor may be used.

An X-ray sensor conventionally used in an article inspection apparatus or the like is generally a scintillator photosensor that generates visible light by incident X-rays, receives the light by a photosensor, and converts it into an electrical signal. The value obtained by integrating the energy represents the density of the image. The X-ray sensors 31 _{1 to} 31 _N used in the electromagnetic wave receiving unit 30 of the article inspection apparatus 20 have X-ray photons transmitted through the object W to be inspected. Each time it is input, it is a photon detection type (CdTe sensor) that outputs a pulse signal having a peak value corresponding to the energy of the photon, and the number of pulses output per unit time represents the density of the image.

  When a photon detection type X-ray sensor is used as described above, if the amount of X-rays input to the X-ray sensor (the number of photons output per unit time) is too large, pulses output from the X-ray sensor. A so-called pile-up phenomenon in which only one peak value (peak value) can be obtained for a plurality of pulse signals due to overlapping of signals occurs, and if this phenomenon occurs with high probability, a correct counting result for each region is obtained. It becomes impossible.

  In order to prevent this, as described above, the amount of X-rays emitted from the electromagnetic wave transmission unit 22 is set to an appropriate range according to the object to be inspected, but if this is still insufficient, for example, the light receiving surface of the X-ray sensor The amount of incident X-rays may be set within an appropriate range using a shielding plate or the like that covers a part of the image. Also regarding these controls, it is assumed that they are registered in the inspection condition registration means 60 as part of the inspection conditions suitable for the inspection object.

The image data generation means 40 outputs signals output from the X-ray sensors 31 _{1 to} 31 _N during a predetermined period (hereinafter referred to as scanning) while the inspection object W passes between the electromagnetic wave transmission unit 22 and the electromagnetic wave reception unit 30. The object to be inspected consists of two-dimensional position information determined by the passing direction of the inspection object W and the arrangement direction of the X-ray sensors and the signal processing result for each position. Are generated for each different wavelength region. The scan time is used to determine a detection unit in the conveyance direction with respect to the inspection object, and is sufficiently short with respect to the article passage time obtained by dividing the length of the inspection object by the conveyance speed.

As described above, the photon detection type X-ray sensors 31 _{1 to} 31 _N output one pulse signal having a peak value corresponding to the energy of the photon with respect to the input of one photon. In addition, since the energy of X-ray photons output from the electromagnetic wave transmitter 22 is not constant and varies, the pulse signals P ₁ and P output from the respective X-ray sensors are accordingly generated as shown in FIG. Variations occur in the peak values H ₁ , H ₂ , H ₃ ,... Of ₂ , P ₃ ,. The peak values having these variations correspond to X-ray wavelengths, respectively.

In other words, X-rays having different energies (corresponding to wavelengths) are mixed, and the peak values H ₁ , H ₂ , H ₃ ,... Of pulse signals output from one X-ray sensor within the scan time. However, it is determined which _{one of} the regions R _{1 to} R _M in which the entire output range of the peak value is divided into a plurality of M (M = 4 in FIG. 2) in advance, and the number of pulse signal inputs within the scan time is determined for each region. If accumulated, a plurality of pieces of image data having different X-ray transmission energy ranges (that is, wavelength regions) can be generated.

In order to realize this, as shown in FIG. 3, the image data generation means 40 converts the output signals of the X-ray sensors 31 _{1 to} 31 _N into digital data by the A / D converters 41 _{1 to} 41 _N , respectively. into a column, it is input to the peak value detection unit ₄₂ 1 through 42 _N.

Each of the peak value detection means 42 _{1 to} 42 _N is for detecting the peak value of the pulse signal from the input data string. For example, the differential value (signal ) Is detected as a peak value of the pulse signal, and the region determination unit 43 ₁ detects the data value in the zero cross swimming as the peak value of the pulse signal. ~ 43 Output to _N.

The area determination means 43 ₁ to 43 _N include boundary value areas L _{1 to} L _M−1 that divide the output range of the peak values into a plurality of M area areas R _{1 to} R _M , and peak value detection means 42 ₁ to 42 _N. 42 _N is compared with the peak value detected at _N , and it is determined which area the peak value falls in, and an area identification signal indicating the area where the peak value enters is output to the accumulation means 44 _{1 to} 44 _N by area. .

Each area accumulating means 44 _{1 to} 44 _N receives the area identification signals output from the area determination means 43 ₁ to 43 _N within the scan time, respectively, and accumulates the number of input area identification signals indicating the same area. The cumulative number for each region within the scan time is obtained and sequentially output.

The cumulative number of area identification signals is the cumulative number of pulse signals having the same peak area among the pulse signals output from one X-ray sensor within the scan time. The accumulated number of area identification signals output from _{1 to} 44 _{N for} each scan time is stored in the image data memory 45 in parallel and in time series, so that X-ray transmission image data for the inspection object for each area is stored. Is obtained.

As a simple example, the scan time is 3 units, the number of X-ray sensors N is 3, the number M of peak values is 3, and the cumulative number of pulse signals is A (order of peak values, rank of scan time, sensor expressed in sequence order) of in the first scan time T1, of the first X-ray pulse signal sensor ₃₁₁ has output, the total number of those whose peak value enters the area R ₁ a (1, 1 , 1), the cumulative number a (2,1,1 those entering the area R _2), the cumulative number to fall region R ₃ and a (3,1,1).

Also, of the second X-ray sensor 31 pulse _{signal 2} is output in the same scan time within T1, the total number of those whose peak value enters the area R ₁ A (1,1,2), in the region R ₂ entering one of the cumulative number a (2,1,2), the cumulative number of those entering the area R ₃ and a (3,1,2).

Further, of the third pulse signal by the X-ray sensor ₃₁₃ is output in the same scan time within T1, the total number of those whose peak value enters the area R ₁ A (1,1,3), in the region R ₂ entering one of the cumulative number a (2,1,3), the cumulative number of those entering the area R ₃ and a (3,1,3).

Similarly, in the next scan time T2, 1 th of the pulse signal X-ray sensor ₃₁₁ has output, the total number of those whose peak value enters the area R ₁ A (1,2,1), region the total number of those entering the R ₂ a (2,2,1), the cumulative number of those entering the area R ₃ and a (3,2,1), the second X-ray sensor ₃₁₂ is a pulse signal output Of these, the cumulative number of those whose peak values fall into the region R ₁ is A (1,2,2), the cumulative number of those whose peak values fall into the region R ₂ is A (2,2,2), and the cumulative number of those whose peak values fall into the region R ₃ Is A (3,2,2), and among the pulse signals output from the _third X-ray sensor 31 ₃ , the cumulative number of the peak values falling within the region R ₁ is A (1,2,3), Assume that the cumulative number of things entering R ₂ is A (2,2,3), and the cumulative number of things entering region R ₃ is A (3,2,3).

Further, in the next scan time T3, of the first X-ray pulse signal sensor ₃₁₁ has output, the total number of those whose peak value enters the region R ₁ A (1, 3, 1), area R the total number of those entering ₂ a (2,3,1), the cumulative number of those entering the area R ₃ and a (3,3,1), of the second X-ray sensor ₃₁₂ is a pulse signal output , the cumulative number that the peak value enters the area R ₁ a (1,3,2), the cumulative number of those entering the area R ₂ a (2,3,2), the cumulative number of those entering the area R ₃ and a (3,3,2), 3 th of pulse signals X-ray sensor ₃₁₃ has output, the total number of those whose peak value enters the region R ₁ a (1,3,3), a region R the total number of those entering ₂ a (2,3,3), the cumulative number of those entering the area R ₃ and a (3,3,3).

From the data obtained in this way, the nine cumulative numbers obtained for the region R ₁ are arranged in the order of scanning time in the horizontal direction and the arrangement order of sensors in the vertical direction as shown in FIG. if arranged in three rows and three columns as the image data of the nine sites of the object by the X-ray energy range corresponding to the region R ₁ (wavelength range) are obtained.

Similarly, if the nine cumulative numbers obtained for the region R ₂ are arranged in 3 rows and 3 columns as shown in FIG. 4B, the X-rays of the inspected object in the energy range corresponding to the region R ₂ can be obtained. If image data is obtained and the nine cumulative numbers obtained for the region R ₃ are arranged in 3 rows and 3 columns as shown in FIG. 4C, the X-ray coverage in the energy range corresponding to the region R ₃ is obtained. Image data of the inspection object is obtained.

  In practice, the number of scans required for foreign object inspection is obtained by dividing the transport time (for example, 0.5 seconds) obtained by dividing the length of the article in the transport direction by the transport speed by the scan time (for example, 1 millisecond). The value (for example, 500) corresponds to the number N of the X-ray sensors (for example, 200).

  When image data for each energy range (wavelength range) corresponding to each peak value region is obtained in this way, the determination means 50 performs conventional subtraction processing on the plurality of image data. By performing predetermined image processing including it, it is possible to determine the presence or absence of foreign matter on the inspection object.

  The method of dividing the peak value region is arbitrary, and as one example, the maximum value of the energy of X-ray photons emitted from the electromagnetic wave transmission unit 22 (in the case of an X-ray tube, acceleration of electrons) What is necessary is just to divide equally between the peak value of the pulse signal which an X-ray sensor outputs with respect to a voltage (theoretical value depending on voltage), and a predetermined reference value (for example, 0). Also, the number of areas can be any number of two or more. First, image data is generated in many areas, and an optimal combination of image data for detecting foreign matter is found for the inspection object. Predetermined image processing including subtraction processing may be performed.

Specifically, for example, assuming that the initial number of regions is 10, image data is generated in each region, and the first region counted from the one with the largest energy is assigned to the region R ₁ described above. An area is assigned to the above-mentioned area R ₂ ,... Is selectively assigned from the initial area to the final area, and predetermined image processing is performed using a plurality of image data in the assigned area. May be. Further, by combining the image data of the first and second regions counted from the larger energy, which was the image data in the above described region R _1, to synthesize the image data of the third and fourth region This is used as the image data of the region R ₂ described above, and the image data of a plurality of initial regions are combined to form the final image data of one region, and a plurality of the combined image data is used. Alternatively, the predetermined image processing may be performed using the synthesized image data and the image data of the initial region that does not include the image data.

  In the above specific example, image data is generated for the initial number of regions, and regions are allocated and image data is synthesized in accordance with a combination of image data optimal for inspection including foreign object detection. When the optimal combination of image data for inspection of an object to be inspected is known, only the image data for the allocated area needs to be generated, and instead of combining a plurality of image data, One image data may be generated by adding the cumulative number of area identification signals. Thereby, the storage area of the image data can be saved.

  Here, the subtraction process will be briefly described. When X-ray transmission data with different energy (wavelength) is obtained for the same part, if the difference process is performed, the influence of the thickness of the part is removed, and the material ( Only the influence of the transmittance) appears, and the difference between the change in the transmittance of the material of the object to be inspected and the change in the transmittance of the material of the foreign matter with respect to the difference in X-ray energy becomes remarkable. Thereby, the detection sensitivity with respect to a foreign material becomes high. In addition to this process, the determination unit 50 performs various filter processes and the like for noise removal and the like to detect foreign matter with higher accuracy.

  Since the plurality of image data obtained by the above method is obtained from the outputs of a plurality of X-ray sensors arranged in a line in a direction orthogonal to the passage direction of the article, compared to the conventional method using two line sensors, Remarkably high-precision image data can be obtained, whereby foreign objects and the like can be detected accurately and can be made compact.

  The determination result of the determination means 50 (good / bad determination signal indicating the presence / absence of a foreign object, a missing item, etc.) is sent to a subsequent sorting device (not shown), and the article determined to be defective is sent from the non-defective path. Will be eliminated.

  In order to enable the inspection apparatus having the above configuration to correctly inspect the inspection object, the energy of the electromagnetic wave output from the electromagnetic wave transmission unit 22, the energy and reception sensitivity of the electromagnetic wave input to the electromagnetic wave reception unit 30, and image data generation The inspection conditions including various parameters necessary for the image data generation processing by the means 40 and various parameters necessary for the foreign object detection processing of the inspection object by the determination means 50 are obtained in advance for each type of the inspection object and stored in the apparatus. Registered in the inspection condition registration means 60, and at the time of product type switching of the object to be inspected, the product type switching means 70 selects the inspection condition of the next inspection target product type from the registered product inspection conditions, and the electromagnetic wave The transmission unit 22, the electromagnetic wave reception unit 30, the image data generation unit 40, and the determination unit 50 are set.

  However, as described above, even if the inspected items of the same variety, the inspected items in which the posture when passing through the passageway is either upright or lying down, for example, frozen meat There is an object to be inspected such that the thickness of the contents in the X-ray transmission direction varies greatly, and for this kind of object to be inspected, it is set in advance by the kind switching means 70 for that kind. When the inspection is performed under the inspection conditions, a correct inspection result may not be obtained due to a difference in passing posture or a difference in thickness.

  In addition, as described above, when materials to be inspected of different varieties are mixed in random order when they are close in terms of material, if they are inspected under the same inspection conditions, the contents will be slightly reduced. There are cases where correct test results cannot be obtained due to differences.

  In order to solve this, in the article inspection apparatus 20 according to the embodiment, the inspection condition registration unit 60 has the same operating conditions for the electromagnetic wave transmission unit 22 and the electromagnetic wave reception unit 30, and the difference between similar types or the same type Optimal inspection conditions are obtained and registered for the processing conditions and determination processing conditions of image data generation that may change depending on the difference in passing posture and the difference in thickness.

  Then, when the electromagnetic wave transmission unit 22 and the electromagnetic wave reception unit 30 have common operating conditions, there is a possibility of being mixed and carried in based on a signal output from the electromagnetic wave reception unit 30 when the inspection object passes through the passage. Each time an object to be inspected passes through the passage, it acquires and registers the feature data that can identify the difference between approximate varieties with the same type, or the difference in passage posture and thickness for the same type. If the feature data is newly acquired, and the newly acquired feature data has been registered and is different from the feature data acquired for the inspection object passed immediately before, the product type identified by the newly acquired feature data, Read the inspection condition for either the passing posture or the thickness, and update the inspection condition including the processing condition of the image data for the inspection object to be inspected with the read inspection condition To have.

  In order to realize this, the inspection condition registration means 60 of the article inspection apparatus 20 according to the embodiment has an inspection condition optimized for a product that may be inspected by the article inspection apparatus 20 as shown in FIG. It is registered.

  The inspection conditions shown in FIG. 5 include the product type N, the operating condition A of the electromagnetic wave transmitting unit 22, the operating condition B of the electromagnetic wave receiving unit 30, the processing condition C of the image data generating unit 40, and the processing condition D of the determining unit 50. .

  Here, the varieties N1 and N2 are examples of varieties that are individually and continuously inspected. For the varieties N1, the optimum operating conditions A1 and B1 and processing conditions C1 and D1 are registered for the varieties. The optimum operating conditions A2 and B2 and processing conditions C2 and D2 are also registered for the product type N2. Note that these inspection conditions are not necessarily different depending on the type, and common inspection conditions may be registered for the types N1 and N2.

  The type N3 is an example of an inspected object that has two passage postures with respect to the passage, and changes irregularly. The identification information of one passage posture (for example, the posture standing upright on the passage) is N3a, The identification information of the other passing posture (for example, the posture hung down on the passage) is N3b. The inspection for the type N3 is performed under the common operating conditions A3 and B3 irrespective of the posture. However, since the amount of X-ray transmission greatly changes depending on the passing posture, image data generation processing suitable for foreign object detection and the like is performed. There is a possibility that the conditions of C3a and C3b change depending on the passing posture. Further, along with the change of the processing condition for generating the image data, there is a possibility that the processing condition suitable for the determination by the determination unit 50 also changes like D3a and D3b. Although there are cases where the determination processing conditions are common, there is a high possibility that the processing state of image data generation is changed in order to perform an accurate inspection.

  The type N4 is an example of an object to be inspected with variations in thickness. Here, the thickness is divided into three levels of large, medium, and small, and information for identifying these is N4a to N4c. Similar to the above, the inspection of the type N4 is performed under the common operating conditions A4 and B4 regardless of the thickness. However, since the amount of X-ray transmission varies greatly depending on the thickness, image data suitable for foreign object detection and the like is detected. The conditions for the generation process may change as C4a to C4c depending on the thickness. In addition, with the change of the processing condition for generating the image data, the processing condition suitable for the determination by the determination unit 50 may change as D4a to D4c.

  The varieties N5 and N6 are similar varieties such as “Meat Gyoza” and “Vegetable Gyoza” described above, and are mixed in random order (including cases where the varieties are switched in a short period of time). These two types are performed under the common operating conditions A5 and B5 regardless of the types, but the X-ray transmission amount changes depending on the contents, so that the image is suitable for detecting foreign matter. There is a possibility that the conditions for data generation processing may change as C5 and C6 suitable for each product type. Further, along with the change of the processing conditions for generating the image data, the processing conditions suitable for the determination by the determination unit 50 may change as D5 and D6 suitable for each product type.

  In the case of normal product type switching for a single inspection product such as the product types N1 and N2, the product type switching means 70 reads the inspection conditions for the product type specified by the operation of the operator, and the electromagnetic wave transmission unit 22, electromagnetic wave transmission It is assumed that the receiving unit 30, the image data generation unit 40, and the determination unit 50 are updated and set.

  In addition, the feature data acquisition means 80 is used in the article inspection apparatus 20 in order to correctly inspect inspected objects that pass through with different types of passing postures or different thicknesses even when similar types are brought together. , Feature data registration means 90 and inspection condition switching means 100 are provided.

  The characteristic data acquisition unit 80 is based on the output signal of the electromagnetic wave receiving unit 30 obtained when the inspection object passes through the passage while the electromagnetic wave transmitting unit 22 and the electromagnetic wave receiving unit 30 are in a predetermined operating condition. Feature data capable of identifying an approximate variety, identifying a passing posture for the same variety, and identifying a thickness is acquired. This predetermined operating condition is basically an operating condition suitable for each product type, but is an operating condition common to the above-mentioned approximate product types carried in mixedly.

  This characteristic data includes the distribution shape of the density distribution of the image of the inspection object, the total density, the area on the image, the change characteristic of the transmittance with respect to the wavelength of the inspection object, the contour shape and area of the contents of the inspection object Etc. Regarding the density distribution of the image of the object to be inspected, a histogram relating to the density of raw image data obtained in a certain wavelength region or image data obtained by difference processing between image data obtained in a plurality of wavelength regions is obtained. Ask. In the case of this processing, a part of the function of the image data generation means 40 may be used.

  An example of the histogram of the image of the inspection object is shown in FIG. The histogram in FIG. 6A is an example in which the object to be inspected is “Gyoza”, and the histogram in FIG. 6B is an example in which the object to be inspected is “serial”. However, the change between the left and right peaks is almost U-shaped in “Gyosa”, while it is almost L-shaped and asymmetric in “Serial”. It differs in that it is changing. In addition, the difference between the left and right peaks is small in “Gyosa”, while there is a large difference in “Serial”. Therefore, it is possible to identify whether the object to be inspected is “Gyosa” or “Serial” from the rough difference in the shape of this histogram. Is possible. Although not shown in the figure, there is a difference in the histogram between the similar varieties `` Gyoza '' and `` Meat Gyoza '' with a high percentage of meat and `` Vegetable Gyoza '' with a high percentage of vegetables. , “Meat Gyoza” and “Vegetable Gyoza” can be identified.

  Moreover, the characteristic of the transmittance (relative value) with respect to the X-ray energy (depending on the wavelength) of the inspection object is shown in FIG. These characteristics are the characteristics of four items, “Gyoza A”, “Gyoza B”, “Serial A”, and “Serial B”, and four items in the region where the X-ray energy is low (wavelength is long). Are separated into distinguishable states. Therefore, if it is known that the object to be inspected is “Gyoza” or “serial”, the transmittance characteristic obtained for the object to be inspected is approximated to any one of FIG. By examining this, it can be determined as one of four items. Further, the product type can be identified from the combination of the transmittance characteristic with respect to the X-ray energy and the shape of the histogram of the density distribution of the image data.

  The above example is an example of identification of approximate varieties (for example, N5 and N6) that are close in terms of material. However, when the passing posture is greatly different for the same varieties as in the case of the varieties N3, a histogram of the density distribution is shown. As shown in FIG. 8, since the histogram of the posture N3b, which has been laid down with respect to the histogram of the upright posture N3a, is greatly shifted to the low density side, the difference in the passing posture can be determined by examining this histogram difference (peak value region determination). Can be identified.

  Similarly, for the type N4 with different thicknesses, the peak positions of the histograms are shifted by the thickness differences N4a to N4c as shown in FIG. Can be identified.

  Further, as described above, not only the shape of the density distribution but also the total density of the histogram (the sum of density × frequency), the area on the density distribution image, the contour shape of the contents of the inspection object obtained from the image data, An area or the like may be included, and any combination thereof may be used.

  In the following description, the feature data P for identifying the type, passage posture, and thickness of the object to be inspected is information on the histogram relating to the density of the image, feature information on the transmittance characteristic, or a combination thereof. And

  There are two main periods for acquiring the feature data, one of which is when acquiring the optimum inspection conditions for new varieties, and the other is as described above. This is a case of obtaining each time the inspected object passes when inspecting an inspected object in which approximate varieties are mixedly carried in, or the inspected object changes in passing posture and thickness even with the same kind.

  When newly registering feature data, the operation conditions of the electromagnetic wave transmission unit 22 and the electromagnetic wave reception unit 30 are set as predetermined operation conditions suitable for the product type, and a sample product to be inspected as a new registration target is assigned to the operator as the electromagnetic wave transmission unit 22. An instruction is given to convey between the electromagnetic wave receiving units 30. For example, in the case of the inspection object N3 in which the passing posture changes, the upright posture and the lying posture are distinguished and conveyed.

  In accordance with this instruction, when the inspection object N3 in the upright posture to be registered with the feature data is conveyed, the characteristic data P3a for the upright posture N3a of the inspection object N3 is acquired, and the inspection condition registration means 60 checks the inspection condition. Is registered in the feature data registration means 90 in association with the posture identification information N3a of the registered product type N3. Further, when the inspection object N3 with the lying posture of the feature data registration target is conveyed, the feature data P3b about the lying posture of the inspection object N3 is acquired, and the inspection condition has been registered in the inspection condition registration means 60. It is registered in the feature data registration means 90 in association with the posture identification information N3b of the product type N3.

  Similarly, with respect to the inspected object N4 whose thickness varies, sample products having different thicknesses are carried in, and feature data P4a to P4c of the inspected object N4 to be registered for each thickness are obtained for each thickness, and the thickness is obtained. It is registered in the feature data registration means 90 in association with the identification information N4a to N4c.

  Then, when the inspection is started for these types, the inspection conditions for the type of inspection target are read in advance by the type switching means 70 and set in each unit. Here, it is assumed that the type, the passing posture, or the thickness of the object to be inspected first is already known, and the inspection conditions corresponding to them are initially set.

  In this way, when the inspection is actually performed in the state where the feature data is registered for the type of the inspection object, the detection condition switching unit 100 allows the inspection object to pass through the passage. Each time the feature data Px (i) acquired by the feature data acquisition means 80 is received, it is determined whether or not the feature data Px (i) has been registered in the feature data registration means 90, and the feature data Px (i) ) Is the same as the feature data Px (i-1) acquired for the inspected object that passed through the passage immediately before, and the newly acquired feature data Px (i) is registered. If it has already been matched with the previous feature data Px (i-1), it is determined that there is no difference in product type, passage posture or thickness, and new feature data Px (i) Against the inspected object That the image data inspection conditions necessary for generation processing or the like is not changed as is, in the same test conditions as the object to be inspected which has been carried just before to perform inspection.

  In addition, when the newly acquired feature data Px (i) has been registered and is different from the previous feature data Px (i-1), a difference in type, a difference in passing posture, or a difference in thickness occurred. The inspection conditions registered in the inspection condition registration means 60 are read out for the product type, passing posture, and thickness identified by the feature data Px (i), and are necessary for image data generation processing and determination processing. The inspection conditions are updated, and image data generation processing and determination processing for the inspection object are performed under the updated inspection conditions. Note that when the inspection conditions are updated, the operation conditions of the electromagnetic wave transmission unit 22 and the electromagnetic wave reception unit 30 are common to the difference in the type, the passing posture, or the thickness. In addition, even if these operating conditions are updated, the operating conditions are not substantially changed.

  As described above, in the article inspection apparatus 20 according to the embodiment, the electromagnetic wave transmission / reception among the inspection conditions including the electromagnetic wave transmission / reception operation conditions and the image data generation / determination processing conditions necessary for the inspection of the inspection object in advance. Inspection conditions including image data processing conditions that may be changed depending on at least one of the type, passage posture, and thickness of the inspection object. Furthermore, based on the received electromagnetic wave output obtained when the inspection object passes through the passage under the common operating conditions for electromagnetic wave transmission and reception, the difference in the type, passage posture, and thickness of the inspection object Any feature data that can be identified is registered in the feature data registration means 90, and each time the inspection object to be inspected passes through the passage, the newly acquired feature data is registered and Inspection When the feature data acquired for the inspection object passed immediately before the target inspection object is different from the feature data acquired, the inspection condition for the type, passage posture, or thickness identified by the newly acquired feature data is read out, The inspection conditions including at least image data generation processing conditions for the inspection object to be inspected are updated with the read inspection conditions.

  For this reason, even if there is a difference in the passing posture and thickness of the same type of inspected objects, or even when mixed with similar types, correct inspection conditions are set for the inspected objects each time. And inspection of goods can be performed accurately.

  The processing for newly registering the inspection condition of the inspection condition registration means 60 will not be described in detail. However, for the products whose posture and thickness change, sample products having different postures and thicknesses can be used under common operating conditions. By carrying it in several times, the article inspection device finds the optimum inspection conditions such as image data generation processing according to the posture and thickness of the product type, and the inspection condition registration means together with information for identifying the product type, posture and thickness 60 is newly registered. In addition, for the approximate product types that are inspected in a mixed manner, the product inspection device can carry out the optimal image data generation process according to the product type by bringing the sample product of each product type several times under common operating conditions. Are newly registered in the inspection condition registration means 60 together with information for identifying the product type. At this time, as described above, feature data that can identify the posture, thickness, or product type is also acquired and newly registered in the feature data registration unit 90.

  In the above embodiment, X-rays are used as electromagnetic waves used for inspecting the inspection object, and a line sensor type X-ray sensor (dual energy X-ray sensor or multi-energy X-ray sensor) is used as the receiving unit. Radiation source 2 sensor dual energy imaging can also be used. In addition to the line sensor type, a high-definition X-ray sensor (an X-ray TDI sensor having an element pitch of 100 μm or less, an X-ray flat panel detector, or the like) can also be used.

  In addition to X-rays, if the electromagnetic wave has an appropriate transmittance for the contents of the object to be inspected or the packaging material and has a wavelength that is safe for food, etc., electromagnetic waves of other wavelengths, for example, It is possible even in the range of ultraviolet light, (far, near) infrared light, microwave, etc. In that case, electromagnetic waves including different wavelength regions are transmitted and received, image data for each wavelength region is generated, and It is possible to perform the inspection including the foreign object detection of the inspection object by the difference processing or the like in the same manner as described above. In the case of using an electromagnetic wave having a wavelength range of “light” such as ultraviolet light and (far, near) infrared light, a photon detection type sensor can be used as in the case of X-rays. In the case of microwaves, although the wavelength range is wide, an electromagnetic wave having a sufficiently narrow wavelength (for example, several mm or less) compared to the width of the object to be inspected is used, and an array antenna for microwaves is used on the receiving unit side. It can respond.

  DESCRIPTION OF SYMBOLS 20 ... Article inspection apparatus, 21 ... Conveyance apparatus, 22 ... Electromagnetic wave transmission part, 30 ... Electromagnetic wave reception part, 40 ... Image data generation means, 45 ... Image data memory, 50 ... Determination means, 60 ... ... Inspection condition registration means 70... Product type switching means 80... Feature data acquisition means 90... Feature data registration means 100.

Claims (5)

An electromagnetic wave transmission unit (22) for outputting electromagnetic waves of different wavelength regions in a passage path through which the inspection object passes;
An electromagnetic wave receiver (30) for receiving an electromagnetic wave output from the electromagnetic wave generator to the passage and transmitted through the object to be inspected;
Image data generating means (40) for generating image data of an object to be inspected with respect to electromagnetic waves in the plurality of different wavelength regions by signal processing on the output of the electromagnetic wave receiving unit;
A determination unit (50) for performing a determination process of the quality of the inspection object from the image data generated by the image data generation unit;
Among the inspection conditions including the operation condition of the electromagnetic wave transmission unit necessary for the inspection of the inspection object, the operation condition of the electromagnetic wave reception unit, the processing condition of the image data of the image data generation unit, and the determination processing condition of the determination unit The operation condition of the electromagnetic wave transmission unit and the electromagnetic wave reception unit is the same, and the image data generation unit may change depending on at least one of the type of inspection object, the passing posture, and the difference in thickness. Inspection conditions including image data processing conditions are pre-registered inspection condition registration means (60);
Based on the output signal of the electromagnetic wave reception unit obtained when the inspection object passes through the passage when the electromagnetic wave transmission unit and the electromagnetic wave reception unit are in the common operating condition, the type of the inspection object, A feature data acquisition means (80) for acquiring feature data capable of identifying either a passing posture or a difference in thickness;
Feature data registration means (90) in which feature data acquired in advance by the feature data acquisition means for the inspection object to be inspected is registered;
Each time the inspection object to be inspected passes through the passage, it receives feature data newly acquired by the feature data acquisition means, and the newly acquired feature data is registered in the feature data registration means and If the feature data acquired for the inspection object passed immediately before is different from the feature data acquired, the inspection condition registration means indicates the inspection condition for any of the product type, the passing posture, and the thickness identified by the newly acquired feature data. And an inspection condition switching means (100) for updating an inspection condition including an image data processing condition of the image data generating means for the inspection object to be inspected with the read inspection condition.   2. The article inspection apparatus according to claim 1, wherein a plurality of different wavelength regions of the electromagnetic wave output by the electromagnetic wave transmitting means are in an X-ray region.   The feature data acquired by the feature data acquisition means includes the shape of the density distribution or the total density of the image of the object to be inspected obtained when the electromagnetic wave transmitting unit and the electromagnetic wave receiving unit are in the common operating condition, or on the image. The article inspection apparatus according to claim 1, wherein at least one of the areas is included.   The characteristic data acquired by the characteristic data acquisition means includes the transmittance for electromagnetic waves in the plurality of wavelength regions of the object to be inspected obtained when the electromagnetic wave transmitting unit and the electromagnetic wave receiving unit are in the common operating condition. The article inspection apparatus according to any one of claims 1 to 3. Output electromagnetic waves of different wavelength regions to the passages through which the inspection object passes, receive electromagnetic waves transmitted through the inspection object, and inspect the electromagnetic waves of different wavelength regions by signal processing on the received output An inspection condition switching method for an article inspection apparatus that generates image data of an object, and performs determination processing of the quality of the inspection object from the generated image data,
Among the inspection conditions including the electromagnetic wave transmission / reception operation conditions necessary for the inspection of the object to be inspected, the image data generation and the determination process processing conditions, the electromagnetic wave transmission / reception operation conditions are common and at least the object to be inspected. Registering the inspection conditions including the processing conditions of the image data that may be changed depending on the type of inspection object, the passing posture, or the difference in thickness;
Based on the received output obtained when the inspected object passes through the passage under the common operating conditions in advance, the type of the inspected object, the passing posture, or the difference in thickness is identified. Registering possible feature data,
Each time the inspection object to be inspected passes through the passage, the characteristic data is newly acquired, and the newly acquired characteristic data is registered and the characteristic data acquired for the inspection object that has passed immediately before If different, the inspection condition for any of the product type, the passing posture, and the thickness identified by the newly acquired feature data is read out, and the image data for the inspection object to be inspected under the read inspection condition And a step of updating the inspection condition including the above processing condition.

Images