Inspection device

Abstract

An inspection device comprising: a light source unit that outputs inspection light; a filter unit that includes an optical element having stepwise filter characteristics in which the transmittance of light changes stepwise at an edge wavelength within a prescribed wavelength region; a photodetection unit that receives, from among incident light from an object, first light that is one of transmitted light transmitted through the optical element and reflected light reflected by the optical element, and second light that includes at least the other of the transmitted light and the reflected light, and outputs luminance data of the first light and luminance data of the second light; a data generation unit that generates characteristic data for the incident light using the luminance data of the transmitted light and the luminance data of the reflected light based on the luminance data of the first light and the luminance data of the second light; and an identification unit that identifies the material of foreign matter on the basis of the characteristic data.

Description

Inspection equipment

[0001] This disclosure relates to an inspection device.

[0002] Conventionally, inspection devices are known that inspect objects by irradiating them with light. For example, one such inspection device is the determination device described in Patent Document 1. This conventional determination device receives reflected light when infrared light is irradiated onto an object. It then calculates correlation information between the spectral intensity of the reflected or absorbed spectrum of the reflected light, excluding the wavelength band between 4.18 μm and 4.42 μm, and spectral data of one or more resin types that have been acquired in advance. The device determines the resin type of the object based on the highest correlation information that is above a preset threshold.

[0003] Japanese Patent Publication No. 2018-100903

[0004] The inspection device described above can be applied, for example, to the inspection of foreign objects in an object. Foreign objects in an object can be made of various materials, and it is desirable to accurately identify foreign objects of various materials.

[0005] This disclosure aims to provide an inspection device and inspection method that can accurately identify the materials of various foreign substances in an object.

[0006] An inspection apparatus in one aspect of the present disclosure is an inspection apparatus comprising: (1) a light source unit that outputs inspection light including a wavelength distribution in a predetermined wavelength range toward an object; a filter unit that includes a first optical element having step-like filter characteristics in which the transmittance of light changes in a step-like manner at a first edge wavelength within the predetermined wavelength range; a light detection unit that receives a first light which includes at least the transmitted light that has passed through the first optical element and the reflected light that has been reflected by the first optical element from the incident light reflected or transmitted by the object, and a second light which includes at least the other of the transmitted light and the reflected light, and outputs luminance data of the first light and luminance data of the second light; a data generation unit that generates characteristic data for the incident light from the object using the luminance data of the transmitted light and the luminance data of the reflected light based on the luminance data of the first light and the luminance data of the second light; and an identification unit that identifies the material of a foreign object in the object based on the characteristic data.

[0007] In the inspection apparatus described in [1] above, a first optical element having the characteristic of its transmittance changing in a step-like manner at a first edge wavelength within a predetermined wavelength range is used to generate characteristic data for incident light from the object based on detected brightness data, which includes brightness data for both transmitted light that has passed through the first optical element and reflected light that has been reflected by the first optical element. The spectrum of the incident light, i.e., the spectral reflectance spectrum of the object, differs depending on the material of the foreign substance in the object. Characteristic data is generated based on the degree of change near the first edge wavelength for such spectral reflectance spectra that differ depending on the material of the foreign substance, and by using the generated characteristic data, the material of the foreign substance in the object can be identified with high accuracy.

[0008] An inspection apparatus in one aspect of this disclosure may be [2] "the inspection apparatus according to [1] above, wherein the first edge wavelength is within the full width at half maximum of the wavelength distribution of the inspection light." According to the inspection apparatus according to [2], characteristic data can be generated based on the degree of change near the central wavelength of the inspection light. In this case, it becomes easier to determine the correspondence between the spectrum of the inspection light and the spectrum of the incident light, making it easier to proceed with the inspection.

[0009] An inspection apparatus in one aspect of the present disclosure may be [3] "the inspection apparatus according to [1] or [2] above, wherein the light detection unit comprises a first image sensor for receiving reflected light and a second image sensor for receiving transmitted light." According to the inspection apparatus described in [3], by simultaneously receiving reflected light and transmitted light with separate image sensors, the apparatus becomes less susceptible to changes in the output of the light source and can accurately detect the brightness of each light.

[0010] An inspection apparatus in one aspect of this disclosure may be [4] "an inspection apparatus according to any one of [1] to [3] above, wherein the first edge wavelength is the central wavelength of the inspection light." According to the inspection apparatus described in [4], the slope component of the spectrum of the incident light near the central wavelength of the inspection light can be reflected in the characteristic data, and the material of the foreign object can be identified based on the difference in the slope component according to the material of the foreign object.

[0011] An inspection apparatus in one aspect of the present disclosure may be [5] "an inspection apparatus according to any one of [1] to [3] above, wherein the first edge wavelength includes a second edge wavelength in which the transmittance of the light changes in a step-like manner between a minimum value and a maximum value, and a third edge wavelength which is a wavelength greater than the second edge wavelength, and in which the transmittance of the light changes in a step-like manner between a maximum value and a minimum value in the opposite direction to the direction of change of the transmittance at the second edge wavelength." According to the inspection apparatus described in [5], the step-like filter characteristics are such that the wavelength region between the second edge wavelength and the third edge wavelength is a transmission region or a cutoff region. For example, by setting the center of the inspection light to the center wavelength of the inspection light, the degree of curvature of the incident light spectrum near the center wavelength of the inspection light can be reflected in the characteristic data, and the material of the foreign object can be identified based on the difference in the degree of curvature according to the material of the foreign object.

[0012] An inspection apparatus according to one aspect of the present disclosure may be [6] "an inspection apparatus according to any one of [1] to [5] above, wherein the first optical element has the step-shaped filter characteristics in each of a plurality of wavelength regions that are different from each other." According to the inspection apparatus according to [6], characteristic data can be obtained corresponding to each of a plurality of wavelength regions that are different from each other. As a result, the material of the foreign object can be identified with even greater accuracy by comparing the characteristic data for each wavelength region.

[0013] An inspection apparatus according to one aspect of the present disclosure may be [7] "an inspection apparatus according to any one of [1] to [6] above, wherein the filter portion further comprises a second optical element having the step-shaped filter characteristics at a fourth edge wavelength having a different wavelength from the first edge wavelength." According to the inspection apparatus according to [7], multiple characteristic data can be generated by combining the luminance data of transmitted light and reflected light generated by the first optical element with the luminance data of transmitted light and reflected light generated by the second optical element, thereby enabling more accurate identification of the material of foreign objects.

[0014] An inspection apparatus relating to one aspect of the present disclosure may be [8] "the inspection apparatus according to [1] or [2] above, wherein the light detection unit is configured such that a plurality of image sensors, including an image sensor that receives transmitted light transmitted through the first optical element by a plurality of pixels arranged in one direction, are arranged along a direction intersecting in that one direction." According to the inspection apparatus described in [8], the light detection unit can be made compact by configuring the light detection unit with an image sensor consisting of a plurality of pixels arranged in one direction. Furthermore, for example, reflected light luminance data can be obtained by the difference between luminance data output from another image sensor among the plurality of image sensors and luminance data of transmitted light output from the image sensor. Furthermore, for example, when an object is transported in a direction intersecting in one direction by a transport unit such as a belt conveyor, the object can be scanned by the inspection apparatus.

[0015] An inspection apparatus according to one aspect of the present disclosure may be [9] "an inspection apparatus according to any one of [1] to [8] above, further comprising a polarization rotation unit arranged in front of the filter unit in the optical path of the incident light and rotating the polarization direction of the incident light." According to the inspection apparatus according to [9], the filter unit exhibits different filter characteristics depending on the polarization direction of the incident light. As a result, multiple characteristic data can be generated, and the material of the foreign object can be identified with even greater accuracy.

[0016] An inspection apparatus according to one aspect of the present disclosure may be

[10] "an inspection apparatus according to any one of [1] to [9] above, further comprising a polarizing plate disposed downstream of the light source in the optical path of the inspection light, wherein the filter unit further includes a polarizing beam splitter that reflects one of the incident light's polarization component parallel to the transmission axis of the polarizing plate and the polarization component perpendicular to the transmission axis of the polarizing plate, and transmits the other of the incident light's polarization component parallel to the transmission axis of the polarizing plate and the polarization component perpendicular to the transmission axis of the polarizing plate, or a polarizing plate that blocks one of the polarization components and transmits the other of the polarization components." According to the inspection apparatus described in

[10] , the polarizing plate and the polarizing beam splitter have a parallel nicol configuration with respect to the polarization component parallel to the polarization direction of the polarizing plate, and a cross-nicol configuration with respect to the polarization component perpendicular to the transmission axis of the polarizing plate. As a result, in parallel nicols, which detect the polarization component parallel to the transmission axis of the polarizer, specular reflected light from transparent objects can be efficiently detected, allowing for accurate determination of whether or not a transparent object is included among the foreign objects. On the other hand, in cross-sectional nicols, which detect the polarization component perpendicular to the transmission axis of the polarizer, specular reflected light can be removed, improving the accuracy of identifying foreign objects other than transparent objects.

[0017] An inspection apparatus in one aspect of the present disclosure may be

[11] "an inspection apparatus according to any one of [1] to

[10] above, wherein the data generation unit generates specular reflection image data based on the characteristic data, and the identification unit determines whether or not the object contains a transparent object based on a comparison between the brightness of the specular reflection image data and a threshold." According to the inspection apparatus described in

[11] , it is possible to determine with high accuracy whether or not the object contains a transparent object by making a determination based on the brightness of the specular reflection light.

[0018] An inspection apparatus according to one aspect of the present disclosure may be

[12] "an inspection apparatus according to any one of [1] to

[11] above, wherein the data generation unit generates, as characteristic data, light intensity ratio image data showing the ratio of the light intensity of transmitted light to the light intensity of reflected light, or light intensity difference image data based on the difference between the light intensity of transmitted light and the light intensity of reflected light, and the identification unit identifies the material of the foreign substance in the object based on either the light intensity ratio image data or the light intensity difference image data." According to the inspection apparatus described in

[12] , since different shapes are observed in the light intensity ratio image data or light intensity difference image data depending on the material of the object, the material of the foreign substance in the object can be identified with even greater accuracy.

[0019] An inspection apparatus relating to one aspect of this disclosure may be

[13] "an inspection apparatus according to any one of [1] to

[12] above, wherein the data generation unit determines an approximate formula for the spectral reflectance spectrum of the object based on the luminance data of the transmitted light and the luminance data of the blocked light, and the identification unit performs identification of the material of the foreign object or analysis of the material composition based on the parameters of the approximate formula." According to the inspection apparatus described in

[13] , the material of the foreign object in the object or analysis of the material composition can be performed with high accuracy based on the parameters of the approximate formula.

[0020] An inspection apparatus in one aspect of this disclosure may be

[14] "the inspection apparatus according to

[13] above, wherein the parameters are parameters based on the slope of a linear approximation formula and the reflectance at the central wavelength." According to the inspection apparatus described in

[14] , the values ​​of the slope and reflectance represent the properties of the material itself, and are therefore effective in accurately performing quantitative analysis of the material.

[0021] An inspection apparatus in one aspect of this disclosure may be

[15] "the inspection apparatus according to

[13] above, wherein the parameters are parameters based on the degree of curvature of a quadratic function approximation and the reflectance at the central wavelength." According to the inspection apparatus according to

[15] , the material identification of foreign matter in an object or the analysis of its material composition can be performed with greater accuracy.

[0022] One aspect of the present disclosure is an inspection method comprising:

[16] "an output step of outputting inspection light including a wavelength distribution in a predetermined wavelength range toward an object; a light guiding step of guiding incident light generated by the reflection or transmission of the inspection light at the object to a first optical element having step-like filter characteristics in which the transmittance of light changes in a step-like manner at a first edge wavelength in the predetermined wavelength range; a light detection step of receiving a first light from the incident light, which includes at least the transmitted light that has passed through the first optical element and the reflected light that has been reflected by the first optical element, and a second light which includes at least the other of the transmitted light and the reflected light, and outputting luminance data of the first light and luminance data of the second light; a generation step of generating characteristic data for the incident light from the object using the luminance data of the transmitted light and the luminance data of the reflected light based on the luminance data of the first light and the luminance data of the second light; and an identification step of identifying the material of a foreign substance in the object based on the characteristic data."

[0023] In the inspection method described in

[16] above, a first optical element having the characteristic of its transmittance changing in a step-like manner at a first edge wavelength within a predetermined wavelength range is used to generate characteristic data for incident light from the object based on detected brightness data, which includes brightness data for both transmitted light that has passed through the first optical element and reflected light that has been reflected by the first optical element. The spectrum of the incident light, i.e., the spectral reflectance spectrum of the object, differs depending on the material of the foreign substance in the object. Characteristic data is generated based on the degree of change near the first edge wavelength for such spectral reflectance spectra that differ depending on the material of the foreign substance, and by using the generated characteristic data, the material of the foreign substance in the object can be identified with high accuracy.

[0024] This disclosure makes it possible to provide an inspection device and inspection method that can accurately identify the materials of various foreign substances in an object.

[0025] Figure 1 shows an example of the configuration of the inspection apparatus according to the first embodiment. Figure 2 shows an example of the irradiation wavelength relative to the scanning position of the object in a sequential lighting method. Figure 3 shows an example of the configuration around the filter section shown in Figure 1. Figure 4 shows an example of the step-shaped filter characteristics of the optical element shown in Figure 1. Figure 5(a) shows an example of the step-shaped filter characteristics in the inspection apparatus according to the first embodiment. Figure 5(b) shows an example of the reflection spectrum of the inspection light. Figure 6 shows an example of the spectral reflection spectrum of the object. Figure 7 shows an example of the step-shaped filter characteristics of the optical element in the inspection apparatus according to the second embodiment. Figure 8(a) shows an example of the step-shaped filter characteristics in the inspection apparatus according to the second embodiment. Figure 8(b) shows an example of the reflection spectrum of the inspection light. Figure 9 shows an example of the spectral reflection spectrum of the object. Figure 10 shows an example of the configuration around the filter section in the inspection apparatus according to the third embodiment. Figure 11(a) shows an example of the step-shaped filter characteristics of the optical element. Figure 11(b) shows an example of the step-shaped filter characteristics of another optical element. Figure 12(a) shows an example of the step filter characteristics in the inspection apparatus according to the third embodiment. Figure 12(b) shows an example of the reflection spectrum of the inspection light. Figure 13 shows an example of the spectral reflection spectrum of the object. Figure 14(a) shows another example of the step filter characteristics in the inspection apparatus according to the third embodiment. Figure 14(b) shows another example of the spectral reflection spectrum of the object. Figure 15(a) shows another example of the step filter characteristics in the inspection apparatus according to the third embodiment. Figure 15(b) shows another example of the spectral reflection spectrum of the object. Figure 16 shows an example of the configuration of the photodetector according to the fourth embodiment. Figure 17(a) shows an example of the configuration of the photodetector according to the fifth embodiment. Figure 17(b) shows an example of the step filter characteristics of an optical element. Figure 17(c) shows an example of the step filter characteristics of another optical element. Figure 18 shows an example of the configuration around the filter section in the inspection apparatus according to the sixth embodiment. Figure 19(a) shows an example of the step filter characteristics of an optical element.Figure 19(b) is a diagram showing an example of the spectral reflectance spectrum of an object. Figure 20 is a flowchart showing an inspection method according to one embodiment of the present disclosure. Figure 21 is a flowchart showing an example of the details of the generation step and identification step shown in Figure 20. Figure 22(a) is a diagram showing an example of light intensity difference image data when a transparent object is included. Figures 22(b) and 22(c) are diagrams showing an example of the relationship between the spectrum of a transparent object and the transmission characteristics of an optical element. Figures 23(a) and 23(b) are diagrams showing an example of the relationship between the spectrum of a transparent object and the transmission characteristics of another optical element. Figures 24(a) and 24(b) are diagrams showing an example of the relationship between the spectrum of a transparent object and the transmission characteristics of yet another optical element. Figure 25 is a flowchart showing an example of an inspection method according to a first modification. Figure 26 is a diagram showing an example of the configuration of an inspection apparatus according to a second modification. Figure 27 is a diagram showing an example of the configuration around the filter section shown in Figure 26. Figures 28(a) and 28(b) are diagrams for conceptually explaining the case where a polarizing beam splitter and a polarizer function as orthogonal nicols. Figures 29(a) and 29(b) are diagrams for conceptually illustrating the case where a polarizing beam splitter and polarizer function as parallel nicols. Figure 30 is a flowchart showing an example of an inspection method when identifying whether or not an object is transparent using the inspection device according to the second modified example. Figure 31 is a diagram showing an example of the configuration of the photodetector according to the third modified example. Figure 32 is a diagram showing an example of the configuration of the inspection device according to the fourth modified example. Figure 33 is a diagram showing an example of the configuration of the inspection device according to the fifth modified example. Figure 34 is a diagram showing an example of the configuration of the filter unit according to the sixth modified example. Figure 35(a) shows the identification result based solely on the wavelength-to-light intensity ratio. Figures 35(b) and 35(c) show the identification result based solely on the relative light intensity value of the central wavelength. Figures 36(a) and 36(b) show the identification result based on the correlation between the relative light intensity difference of the central wavelength and the wavelength-to-light intensity ratio. Figure 37 shows the identification result based on the correlation between the relative light intensity difference of the central wavelength and the relative light intensity difference of another central wavelength. Figure 38 is a graph showing an example of the reflection spectral characteristics of inspection light irradiated onto an object. Figure 39 is a graph showing an example of the step filter characteristics of an optical element. Figure 40 is a graph showing an example of the bandpass filter characteristics of an optical element.Figure 41 is a graph showing an example of a derived approximation formula.

[0026] The embodiments of this disclosure will be described in detail below with reference to the drawings. In each drawing, the same or corresponding parts are denoted by the same reference numerals, and redundant descriptions are omitted. [First Embodiment]

[0027] Figure 1 is a schematic diagram showing the configuration of an inspection apparatus according to one embodiment of the present disclosure. The inspection apparatus 1 shown in Figure 1 is configured as an apparatus for inspecting foreign matter W in an object S by irradiation with inspection light L1. As will be described later, the inspection apparatus 1 uses an optical element 40 having a step-like filter characteristic in which the transmittance of light changes in a step-like manner between a maximum value and a minimum value at the edge wavelength within a predetermined wavelength range. In the inspection apparatus 1, incident light L2 generated by the reflection of inspection light L1 from the object S is incident on the optical element 40. The inspection apparatus 1 receives transmitted light Lt that has passed through the optical element 40 and reflected light Lr that has been reflected by the optical element 40 from the incident light L2. The reflected light Lr may be included in the remaining light of the incident light L2 other than the transmitted light Lt. The inspection apparatus 1 generates light intensity difference image data GC of the object S based on the luminance data of the transmitted light Lt and the luminance data of the reflected light Lr, and identifies the material of the foreign matter W in the object S based on the generated light intensity difference image data GC.

[0028] Examples of the object S include food products. Examples of food products include beef, pork, chicken, lamb, or processed foods made from these. The object S is not limited to food products, but may also include other items such as electronic components. Examples of foreign matter W in the object S include resin pieces of various materials. Furthermore, foreign matter W is not limited to those that are opaque to visible light, but may also include those that are transparent to visible light, such as glass pieces, film pieces, and vinyl pieces.

[0029] As shown in Figure 1, the inspection device 1 comprises a transport unit 2, a light source unit 3, a filter unit 4, a light detection unit 5, a data generation unit 6, an identification unit 7, and a display unit 8. The transport unit 2 is the part that transports the object S in a predetermined direction. The transport unit 2 is, for example, a belt conveyor. The transport unit 2 transports the object S horizontally at a constant speed toward the irradiation position of the inspection light L1 from the light source unit 3. As a result, the object S is scanned relative to the irradiation position of the inspection light L1.

[0030] The light source unit 3 outputs inspection light L1 containing a wavelength distribution in a predetermined wavelength range toward the object S. In this embodiment, the light source unit 3 includes light source devices 31 and 32 that can output inspection light L1 toward the object S from two directions, the upstream and downstream sides in the transport direction. The inspection light L1 is, for example, a directional light beam. Examples of light source devices 31 and 32 that output such light include multiband LED bar lighting devices. Such lighting devices include, for example, a light source device with a center wavelength of λ １ = 1200 nm LED, center wavelength λ ２ = 1300 nm LED, center wavelength λ ３ = 1450 nm LED, center wavelength λ ４ = 1550 nm LED, center wavelength λ ５ = 1650 nm LEDs are incorporated in an alternating pattern. By including two light source devices 31 and 32 in the light source unit 3, for example, shadows that depend on the direction of illumination are less likely to occur, and uniform illumination becomes possible.

[0031] The light source devices 31 and 32 may consist of, for example, a combination of a white light source capable of irradiating the surface of an object S in a line and a filter switching mechanism. The white light source is a light source that outputs light with a broad wavelength spectrum, such as an SC (Super Continuum) light source, a white LED light source, or a white lamp light source. The filter switching mechanism includes, for example, a plurality of bandpass filters and a switching member (wheel, slider, etc.) that holds the plurality of bandpass filters in a switchable manner. The switching member may switch the plurality of bandpass filters on the optical axis of the light output from the white light source. The plurality of bandpass filters are filter members that transmit light in different wavelength regions from the light irradiated by the white light source. With this configuration, the light emitted from the white light source is irradiated onto the object S after passing through the bandpass filters set by the filter switching mechanism. In this case, by rapidly switching the filter switching mechanism to a predetermined bandpass filter, the light source unit 3 irradiates the object S with inspection light L1 having a wavelength distribution in a predetermined wavelength region corresponding to the set bandpass filter. Furthermore, the light source devices 31 and 32 may be composed of a tunable filter using liquid crystal instead of the filter switching mechanism described above. Also, the light source devices 31 and 32 may be composed of a tunable laser instead of the white light source described above. In this case, the filter switching mechanism may not be necessary.

[0032] From the light source unit 3, inspection light L1 of any wavelength is rapidly switched and output toward the object S. Each LED constituting the light source unit 3 lights up sequentially in synchronization with the frame rate of the image sensor described later. In this case, as shown in Figure 2, for the object S being transported at a constant speed by the transport unit 2, the center wavelength is λ １ ~λ ５Inspection light L1 in each wavelength range is repeatedly irradiated in sequence. The light source unit 3 may be composed of a tunable light source or a white light source equipped with multiple bandpass filters. In the inspection device 1, lenses LZ1 may be provided in the optical path of the inspection light L1 between the light source unit 31 and the object S, and between the light source unit 32 and the object S. Lens LZ1 focuses the inspection light L1 output from the light source units 31 and 32 onto the object S. Lens LZ1 is, for example, a cylindrical lens. In the inspection device 1, lens LZ2 may be provided in the optical path of the incident light L2 between the object S and the filter unit 4. Lens LZ2 focuses the incident light L2 from the object S onto the filter unit 4. Lens LZ2 is, for example, a camera imaging lens.

[0033] As shown in Figure 3, the filter section 4 includes an optical element 40, a prism 41, and a prism 42. The prisms 41 and 42 are arranged to overlap each other in this order along the incident direction of the incident light L2.

[0034] The prism 41 has a roughly triangular prism shape with three sides: an input surface 41a perpendicular to the direction of incidence of the incident light L2, a separation surface 41b inclined from the surface perpendicular to the direction of incidence of the incident light L2, and an output surface 41c connecting the input surface 41a and the separation surface 41b. The prism 42 is optically coupled to the prism 41 via the separation surface 41b. The prism 42 has a roughly rectangular prism shape with an input surface 42a adjacent to the separation surface 41b, and an output surface 42c facing the input surface 42a and substantially perpendicular to the direction of incidence of the incident light L2 as its side.

[0035] The optical element 40 (first optical element) is formed on the separation surface 41b. The optical element 40 has a step-like filter characteristic in which the transmittance of light changes in a step-like manner between a maximum value and a minimum value at the edge wavelength (first edge wavelength) within a predetermined wavelength region including the central wavelength of the inspection light L1. Specifically, the optical element 40 is composed of a filter having a step-like filter characteristic. For example, the optical element 40 is a dichroic mirror. The dichroic mirror is composed of, for example, a dielectric multilayer film.

[0036] FIG. 4 is a diagram showing an example of the step filter characteristics. The optical element 40 has step filter characteristics in each of a plurality of different wavelength regions. As shown in FIG. 4, in the optical element 40, the transmittance changes stepwise between the maximum value and the minimum value at each edge wavelength so as to correspond to each of the central wavelengths λ １ to λ ５ . The edge wavelength is the wavelength at which the transmittance changes between the minimum value and the maximum value, and may be referred to as the cut-off wavelength. The edge wavelength exists within a predetermined wavelength region centered on the central wavelengths λ １ to λ ５ . As shown in FIG. 4, the edge wavelength exists within the full width at half maximum Δλ ｗ of the wavelength distribution of the inspection light L1. In the example of FIG. 4, there are five edge wavelengths that coincide with each of the central wavelengths λ １ to λ ５ . The step filter characteristics exhibit multi-edge filter characteristics showing the characteristics of an edge filter at each of the central wavelengths λ １ to λ ５ . In the example of FIG. 4, the minimum value of the transmittance is, for example, 0%, and the maximum value of the transmittance is, for example, 100%. The transmittance and the reflectance of the optical element 40 have a negative correlation, and when one increases, the other decreases. The maximum value of the transmittance may be 90% or more and 100% or less. Also, the minimum value of the transmittance may be 0% or more and 10% or less.

[0037] In the example of FIG. 4, in the wavelength region smaller than the central wavelength λ １ , the transmittance is constant near 0%. The transmittance increases from 0% to 100% at the central wavelength λ １ of the inspection light L1. In the wavelength region between the central wavelength λ １ and the central wavelength λ ２ , the transmittance is constant near 100%. The transmittance decreases from 100% to 0% at the central wavelength λ ２ . In the wavelength region between the central wavelength λ ２ and the central wavelength λ ３ , the transmittance is constant near 0%. The transmittance increases again from 0% to 100% at the central wavelength λ ３ . The transmittance increases from 0% to 100% again at the central wavelength λ ３and central wavelength λ ４ In the wavelength range between λ, the transmittance is constant at around 100%. (Center wavelength λ) ４ At this point, the transmittance decreases from 100% to 0%. (Center wavelength λ) ４ and central wavelength λ ５ In the wavelength range between λ, the transmittance is constant at around 0%. (Center wavelength λ) ５ At this point, the transmittance increases again from 0% to 100%. (Center wavelength λ) ５ In the wavelength range greater than this, the transmittance remains constant at around 100%. In the following explanation, the region where the transmittance is at its maximum value is sometimes referred to as the transmittance region, and the region where the transmittance is at its minimum value is sometimes referred to as the cutoff region. Note that an increase in transmittance at a wavelength means that the transmittance increases as the wavelength increases from that wavelength, and a decrease in transmittance at a wavelength means that the transmittance decreases as the wavelength increases from that wavelength (the same applies below).

[0038] Let's consider a case where the optical element 40 has filter characteristics such that wavelength regions where transmittance increases and wavelength regions where transmittance decreases alternately. In this case, as shown in Figure 4, the step-like filter characteristics become multi-edge filter characteristics, which makes it possible to narrow the distance between adjacent wavelength regions. In the example in Figure 4, the center wavelength λ １ ~λ ５ The minimum interval between each wavelength is 100 nm. As a result, the characteristics are compatible between adjacent wavelength regions. For example, the central wavelength λ １ The transmission region and center wavelength λ of a long-pass filter with edge wavelength λ ２ The transmission bandwidth of the short-pass filter with edge wavelength is composed of the same wavelength region. Let's assume that the optical element 40 does not have a step filter characteristic, but rather a filter characteristic in which wavelength regions where transmittance gradually increases and wavelength regions where transmittance gradually decreases are alternately repeated. In this case, it may be necessary to sufficiently widen the distance between adjacent wavelength regions. From the viewpoint of design freedom for the optical element 40 as described above, it can be said that giving it a step filter characteristic is effective in improving the degree of freedom.

[0039] Refer to Figure 3 again. The prisms 41 and 42 of the filter section 4 function as follows. Incident light L2 is incident on prism 41 from the input surface 41a. The prism 41 then transmits the incident light L2, which has passed through the input surface 41a and the inside of the medium of the prism 41, in the wavelength region where the transmittance of the step-shaped filter characteristics of the optical element 40 formed on the separation surface 41b is at its maximum value. The transmitted light passes through the separation surface 41b and becomes transmitted light Lt. At the same time, the prism 41 reflects the incident light L2 in the wavelength region where the transmittance of the step-shaped filter characteristics is at its minimum value. The reflected light is reflected light Lr reflected by the optical element 40. After being reflected by the optical element 40, the reflected light Lr is further totally reflected at the input surface 41a and output to the outside from the output surface 41c. In the prism 42, transmitted light Lt is received at the input surface 42a, and the transmitted light Lt that has passed through the input surface 42a and the inside of the medium of the prism 42 is output to the outside from the output surface 42c.

[0040] The light detection unit 5 receives a first light from the incident light L2, which is either the transmitted light Lt that has passed through the optical element 40 or the reflected light Lr that has been reflected by the optical element 40, and a second light that includes at least the other of the transmitted light Lt and the reflected light Lr. In the example in Figure 3, the transmitted light Lt that has passed through the optical element 40 from the incident light L2 is received as the first light, and the reflected light Lr that has been reflected by the optical element 40 is received as the second light. As shown in Figure 3, the light detection unit 5 has an image sensor (first image sensor) 51 that detects the reflected light Lr and an image sensor (second image sensor) 52 that detects the transmitted light Lt. The image sensor 51 is positioned so that its light-receiving surface is close to the output surface 41c of the prism 41, and the image sensor 52 is positioned so that its light-receiving surface is close to the output surface 42c of the prism 42. The image sensors 51 and 52 have a plurality of pixels arranged in one or two dimensions on their respective light-receiving surfaces. Image sensor 51 outputs luminance data based on the luminance of the received reflected light Lr to the data generation unit 6. Image sensor 52 outputs luminance data based on the luminance of the received transmitted light Lt to the data generation unit 6.

[0041] The image sensors 51 and 52 are, for example, line sensors. A line sensor is, for example, an image sensor having multiple pixels arranged in one dimension. When imaging is performed using the line sensors while moving the object S, a two-dimensional image can be created by continuously stitching together the images captured in one dimension using the line sensors in the data generation unit 6. In this case, if multiple light sources of different wavelengths are turned on sequentially, the data generation unit 6 may stitch together images captured at the same wavelength.

[0042] The image sensors 51 and 52 may be, for example, area sensors. An area sensor is, for example, an image sensor having multiple pixels arranged in two dimensions. When imaging is performed using the area sensor while moving the object S, the data generation unit 6 continuously adds the two-dimensional images that have been shifted in accordance with the movement of the object S to stitch the images together. In this case, by increasing the overlap of the images, higher sensitivity and a higher dynamic range can be obtained than when using a line sensor. When multiple light sources of different wavelengths are turned on sequentially, the data generation unit 6 may stitch together images captured at the same wavelength.

[0043] The image sensors 51 and 52 may be, for example, TDI (Time Delay Integration) sensors. In a TDI sensor, as with a CCD sensor, vertical transfer is performed in units of one line during charge readout. By synchronizing the timing of the transfer with the timing of the movement of the image of the object S incident on the light-receiving surface of the TDI sensor, it is possible to expose the sensor for the same number of vertical stages as a CCD sensor. This makes it possible to image moving objects at high speed and with high sensitivity.

[0044] A bandpass filter (not shown) may be placed upstream of the image sensors 51 and 52. Such a bandpass filter may be, for example, a filter that removes light from wavelengths other than the predetermined wavelength range (a predetermined wavelength range including the central wavelength of the inspection light L1) described above. The image sensors 51 and 52 may be directly bonded to the filter section 4, bonded via mechanical components, or arranged spatially separated from the filter section 4.

[0045] Returning to Figure 1, the data generation unit 6, identification unit 7, and display unit 8 are physically composed of a computer 10 equipped with memory devices such as RAM and ROM, a processor (arithmetic circuit) such as a CPU, and a communication interface. For example, a personal computer, a cloud server, or a smart device (smartphone, tablet terminal, etc.) can be used as the computer 10. The computer 10 may be composed of an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit). The computer 10 is connected to the light detection unit 5 via wired or wireless means so as to be able to communicate information.

[0046] The data generation unit 6 generates characteristic data for inspecting the object S based on the light reception results from the light detection unit 5. In this embodiment, the data generation unit 6 generates characteristic data for incident light L2 from the object S based on the luminance data of transmitted light Lt and the luminance data of reflected light Lr. In this disclosure, the data generation unit 6 generates light intensity ratio image data GL and light intensity difference image data GC of the object S as characteristic data. The data generation unit 6 classifies each of the multiple luminance data acquired at different center wavelengths for each line by wavelength region and synthesizes the luminance data for each wavelength region. Then, the data generation unit 6 generates light intensity ratio image data GL or light intensity difference image data GC for each wavelength region by performing the following calculations on the luminance data obtained by synthesis. The data generation unit 6 outputs the generated light intensity ratio image data GL or light intensity difference image data GC to the identification unit 7.

[0047] The identification unit 7 identifies the material of the foreign object W in the object S based on characteristic data. The identification unit 7 identifies the material of the foreign object W in the object S based, for example, on either the light intensity ratio image data GL or the light intensity difference image data GC. After identifying the material of the foreign object W, the identification unit 7 outputs information indicating the identification result to the display unit 8. The display unit 8 displays the inspection result of the object S. The display unit 8 is composed of, for example, a general display, a touch panel display, etc. The display unit 8 displays the identification result of the material of the foreign object W based on the information indicating the identification result output from the identification unit 7. There are no particular restrictions on the display method, and in addition to displaying the identification result of the material of the foreign object W, the light intensity ratio image data GL or the light intensity difference image data GC may be displayed together with the identification result.

[0048] The light intensity ratio image data GL is data that shows the ratio of the light intensity of transmitted light Lt to the light intensity of reflected light Lr. The pixel value of each pixel included in the light intensity ratio image data GL is the light intensity, which is the luminance value of the reflected light Lr received by each pixel of the image sensor 51. Ｒ The amount of light, which is the brightness of the transmitted light Lt received by each pixel of the image sensor 52, is I Ｔ In that case, (I Ｔ / I Ｒ The light intensity difference image data GC is data based on the difference between the light intensity of transmitted light Lt and the light intensity of reflected light Lr. In this disclosure, the data generation unit 6 generates data that shows the relative value of the light intensity difference, which is the relative value of the light intensity difference to twice the sum of the light intensities, as data based on the light intensity difference. The pixel value of each pixel included in the light intensity difference image data GC is the light intensity of reflected light Lr I Ｒ and the amount of transmitted light Lt I Ｔ From, (I Ｔ -I Ｒ ) / (2 × (I Ｔ +I Ｒ It is represented as )).

[0049] Here, the light intensity ratio is (I Ｔ / I Ｒ ), and the relative value of the difference in light intensity (I Ｔ -I Ｒ ) / (2 × (I Ｔ +I ＲLet's explain this in more detail. For the sake of explanation, we will assume the step filter characteristics shown in Figure 5(a) and the reflection spectrum of the inspection light L1 shown in Figure 5(b). Figures 5(a) and 5(b) show the center wavelength λ shown in Figure 4. １ , center wavelength λ ３ , and the central wavelength λ ５ The characteristics may correspond to any of the following. As shown in Figure 5(a), the transmittance of the filter section 4 is the center wavelength λ Ｍ The value is increased from 0% to 100% with respect to the edge wavelength. Here, the center wavelength λ Ｍ Smaller than λ Ａ Then the transmittance is 0%, and the central wavelength λ Ｍ Larger than λ Ｂ Then the transmittance is 100%. As shown in Figure 5(b), in the reflection spectrum of the inspection light L1, the center wavelength λ Ｍ At this point, the intensity of the inspection light L1 is at its maximum, and the wavelength λ Ａ and wavelength λ Ｂ Then the intensity of the inspection light L1 becomes zero. Here, the reflection spectrum of the inspection light L1 is the center wavelength λ Ｍ We assume that the waveform follows a normal distribution centered at [a specific point]. The reflection spectrum of the inspection light L1 is, for example, the spectrum obtained from a standard white calibration plate used for calibrating the light source when the light source devices 31 and 32 are multiband LED bar lighting devices. In the following explanation, the reflection spectrum of the inspection light L1 will be denoted as the function j(λ).

[0050] The spectral reflectance spectrum, which is the spectral characteristics of the reflected light from the object S, is a different spectrum depending on the object S being inspected, and an example is shown in Figure 6. As shown in Figure 6, the spectral reflectance spectrum of the object S is, wavelength λ Ａ from wavelength λ Ｂ It exhibits a nonlinear characteristic in which the reflectance r gradually increases over time. In the following explanation, the spectral reflectance spectrum of the object S will be denoted as the function r(λ). Furthermore, in the following explanation, the approximation r of the function r(λ) will be used. ０ (λ) may be used. In the example in Figure 6, the approximation r ０ (λ) is a linear approximation and is expressed by equation (1). ０ (λ) = k(λ - λ) Ｍ ) + rＭ ... (1) In equation (1), r Ｍ The central wavelength is λ Ｍ This is the reflectance at the center wavelength λ. Ｍ This is the rate of change of the spectral reflectance spectrum with respect to wavelength in the vicinity, or in other words, the central wavelength λ Ｍ This is the slope of the spectral reflectance spectrum in the vicinity. k is expressed by equation (2): k = Δr / Δλ ... (2)

[0051] Intensity of reflected light Lr I Ｒ , transmitted light Lt light intensity I Ｔ , and the amount of incident light L2 I ２ This is calculated by the following formulas (3) to (6). As shown in equations (3) and (4), the amount of light I Ｒ and light intensity I Ｔ This is expressed as the integral of the product of the function j(λ) of the inspection light L1 and the function r(λ) of the object S. As shown in equation (5), the amount of light I of the incident light L2 ２ is, light intensity I Ｒ and light intensity I Ｔ This is the sum of the two values. In equation (5), R is an intrinsic value of the object S, and Jo is calculated by equation (6).

[0052] Next, by substituting equation (1) into equations (3) and (4), equations (7) and (8) are obtained.

[0053] Next, the following transformation formulas (formulas (9) to (14)) are used to replace the integral expressions in equations (7) and (8) with constants. In the above conversion formula, the conversion value J Ｒ０ , J Ｔ０ This can be determined from the spectrum obtained from a standard white calibration plate for each wavelength of LED when the light source devices 31 and 32 are multiband LED bar lighting devices. Conversion value J Ｒ１ , J Ｔ１ This is the wavelength data measured from the reflection spectrum data of the previously measured test light L1 and the converted value J. Ｒ０ , J Ｔ０ It can be calculated numerically using and .

[0054] By substituting the above conversion formula into formulas (7) and (8), simplified formulas (15) and (16) of formulas (7) and (8) are derived.

[0055] Subsequently, from formulas (15) and (16), the light quantity ratio (I Ｔ / I Ｒ ), and the relative value of the light quantity difference ((I Ｔ - I Ｒ ) / (2×(I Ｔ + I Ｒ )) are calculated by the following formulas (17) and (18). Note that the slope k and the reflectance r Ｍ are calculated by the following formulas (19) to (21). As shown in formulas (17) and (18), values other than (k / r Ｍ ) can be treated as constants determined by the reflection spectrum of the inspection light L1 and the characteristics of the dichroic mirror of the optical element 40. That is, they can be said to be known values. Therefore, since the changes in the light quantity ratio and the relative value of the light quantity difference are brought about only by the value (k / r Ｍ ) that depends on the substance of the object S, the material can be discriminated from the value of (I Ｔ / I Ｒ ), or ((I Ｔ - I Ｒ ) / (2×(I Ｔ + I Ｒ )). Also, by obtaining in advance the values of the conversion values J Ｒ１ , J Ｔ１ , the conversion values J Ｒ０ , J Ｔ０ , and J １ , J ０ based on these conversion values through actual measurement of the light source spectrum, the values of the slope k and the reflectance r Ｔ of the measured transmitted light Lt and the light quantity I Ｒ of the reflected light Lr can be obtained. Since these are values representing the characteristics of the material itself, they are effective when characteristics such as sugar content, fat content, moisture content, protein, pH, etc. are required in the quantitative analysis of the material. The slope k and the reflectance r Ｍ ​​Since the value represents the material's properties themselves, it is effective for accurately performing quantitative analysis of materials. The target of quantitative analysis may be the material's composition itself, or it may be an evaluation index for its functional aspects when used (for example, numerical values ​​for properties such as sugar, fat, moisture, protein, and pH, or the freshness, sugar content, and acidity of food).

[0056] In the inspection device 1 described above, an optical element 40 having the characteristic of its transmittance changing in a step-like manner at edge wavelengths within a predetermined wavelength range is used to generate characteristic data for incident light L2 from the object S based on the luminance data of the transmitted light Lt that has passed through the optical element 40 and the reflected light Lr that has been reflected by the optical element 40. The spectrum of the incident light L2, that is, the spectral reflectance spectrum of the object S, differs depending on the material of the foreign matter W in the object S. Characteristic data is generated based on the degree of change near the first edge wavelength for such spectral reflectance spectra that differ depending on the material of the foreign matter W, and by using the generated characteristic data, the material of the foreign matter W in the object S can be identified with high accuracy.

[0057] The edge wavelength is the full width at half maximum (Δλ) of the wavelength distribution of the inspection light L1. ｗ It exists within. According to this, characteristic data can be generated based on the degree of change near the central wavelength of the inspection light L1. In this case, it becomes easier to determine the correspondence between the spectrum of the inspection light L1 and the spectrum of the incident light L2, making it easier to proceed with the inspection.

[0058] The light detection unit 5 includes an image sensor 51 that receives reflected light Lr and an image sensor 52 that receives transmitted light Lt. By simultaneously receiving reflected light Lr and transmitted light Lt with separate image sensors, the unit becomes less susceptible to changes in the output of the light source, and the brightness of each light can be accurately detected.

[0059] The edge wavelength of the optical element 40 of the inspection device 1 is the central wavelength of the inspection light L1. This allows the slope component of the spectrum of the incident light L2 near the central wavelength of the inspection light L1 to be reflected in the characteristic data, and the material of the foreign object W can be identified based on the difference in the slope component according to the material of the foreign object W.

[0060] The optical element 40 has step-like filter characteristics in each of several different wavelength regions. This allows for obtaining characteristic data corresponding to each of the multiple different wavelength regions. As a result, by comparing the characteristic data for each wavelength region, the material of the foreign object W can be identified with even greater accuracy.

[0061] The data generation unit 6 generates characteristic data, namely, a light intensity ratio image data GL showing the ratio of the light intensity of transmitted light Lt to the light intensity of reflected light Lr, and a light intensity difference image data GC based on the difference between the light intensity of transmitted light Lt and the light intensity of reflected light Lr. The identification unit 7 identifies the material of the foreign object W in the object S based on either the light intensity ratio image data GL or the light intensity difference image data GC. As a result, different trends are observed in the light intensity ratio image data GL or the light intensity difference image data GC depending on the material of the object S, making it possible to identify the material of the foreign object W in the object S with even greater accuracy. [Second Embodiment]

[0062] The optical element 40 may have characteristics other than multi-edge filter characteristics as a step-type filter characteristic. For example, the edge wavelength corresponding to each center wavelength (first edge wavelength) includes a second edge wavelength in which the transmittance of light changes in a step-like manner between a minimum and a maximum value, and a third edge wavelength which is a wavelength larger than the second edge wavelength in which the transmittance of light changes in a step-like manner between a maximum and a minimum value, in the opposite direction to the direction of the change in transmittance at the second edge wavelength. The inspection apparatus 1A according to the second embodiment differs from the inspection apparatus 1 in that the optical element 40 has multi-bandpass filter characteristics instead of multi-edge filter characteristics, and the method of generating the light intensity ratio image data GL and light intensity difference image data GC in the data generation unit 6 is different. As shown in Figure 7, the optical element 40 in the inspection apparatus 1A according to the second embodiment has a step-type filter characteristic in the range Δλ of a predetermined wavelength region. ａ The filter may have multibandpass characteristics that include a transmission region within which the transmittance of transmitted light Lt is maximized. The range here is Δλ. ａ This refers to the central wavelength λ １ ~λ ５The range is approximately 20 nm to 50 nm, centered on each of these points. ａ For example, the full width at half maximum Δλ ｗ It is smaller than. In the multibandpass filter characteristics, the edge wavelength corresponding to each center wavelength includes an edge wavelength (second edge wavelength) at which the light transmittance changes in a step-like manner from minimum to maximum value, and a third edge wavelength that is larger than the second edge wavelength and at which the light transmittance changes in a step-like manner from maximum to minimum value. In the example in Figure 7, for example, the center wavelength λ １ The corresponding edge wavelength is the center wavelength λ as the second edge wavelength. １ Rather than Δλ ａ / 2 Includes smaller wavelengths, with the center wavelength λ as the third edge wavelength. １ Rather than Δλ ａ It includes wavelengths that are 2 times larger. In other words, the multibandpass filter characteristics are such that the center wavelength λ １ ~λ ５ The range Δλ centered on each of these points. ａ This region is designated as the transmission region, and the other wavelength regions are designated as the blocking region.

[0063] Details of the light intensity ratio image data GL and light intensity difference image data GC in the data generation unit 6 of the inspection device 1A will be explained. For the purposes of this explanation, we will assume the step filter characteristics shown in Figure 8(a), the reflectance spectrum of the inspection light L1 shown in Figure 8(b), and the spectral reflectance spectrum of the object S shown in Figure 9. Figures 8(a), 8(b), and 9 show the center wavelength λ shown in Figure 7. １ ~λ ５ The characteristics may correspond to any of the following. As shown in Figure 8(a), the filter section 4 has a center wavelength λ Ｍ Wavelength (λ) centered on Ｍ -Δ) to wavelength (λ Ｍ The range of +Δ) is the range Δλ ａ This corresponds to the transmission band. Here, the wavelength (λ) Ｍ λ smaller than -Δ) Ａ Then the transmittance is 0%, and the wavelength (λ Ｍ λ greater than +Δ) ＢTherefore, the transmittance is 0%. The spectral reflectance spectrum shown in Figure 8(b) is the same as the spectral reflectance spectrum shown in Figure 5(b).

[0064] As shown in Figure 9, the spectral reflectance spectrum of the object S is, wavelength λ Ａ From the central wavelength λ Ｍ The reflectivity gradually decreases as the wavelength λ approaches its center. Ｍ from λ Ｂ It exhibits a nonlinear characteristic where the reflectance gradually increases over time. In the example in Figure 9, the approximate r ０ (λ) is a quadratic function approximation and is expressed by equation (22). ０ (λ) = a(λ - λ) Ｍ ) ２ +r Ｍ ... (22) In equation (22), a is an approximation of r ０ This is a quantity that represents the degree of curvature of (λ). For example, when a is positive, the curve of the approximation r0(λ) is concave downwards, and when a is negative, r ０ The curve (λ) is concave upwards. Also, the larger the absolute value of a, the more the approximation r ０ The curve of (λ) becomes steeper, and the smaller the absolute value of a, the closer the approximation r is. ０ The curve of (λ) becomes flatter.

[0065] Intensity of reflected light Lr I Ｒ , transmitted light Lt light intensity I Ｔ This is expressed by equations (23) and (24) below.

[0066] Next, by substituting equation (22) into equations (23) and (24), equations (25) and (26) are obtained.

[0067] Next, the following transformation formulas (formulas (27) to (35)) are used to replace the integral expressions in equations (25) and (26) with constants. In the above conversion formula, the conversion value J Ｒ０ , J Ｔ０ This can be determined from the spectra obtained from a standard white calibration plate for each wavelength of LED when the light source devices 31 and 32 are multiband LED bar lighting devices. Conversion value J Ｒ１ , JＴ１ , J Ｒ２ , J Ｔ２ This is the wavelength data measured from the reflection spectrum data of the previously measured test light L1 and the converted value J. Ｒ０ , J Ｔ０ It can be calculated numerically using and .

[0068] By substituting the above conversion formulas into equations (25) and (26), simplified equations (36) and (37) are derived from equations (25) and (26).

[0069] Next, from equations (36) and (37), the light intensity ratio is (I Ｔ / I Ｒ ), and the relative value of the difference in light intensity (I Ｔ -I Ｒ ) / (2 × (I Ｔ +I Ｒ )) is calculated by the following formulas (38) and (39). Note that the degree of curvature a and reflectance r are also considered. Ｍ This is calculated by the following equations (40) to (42). As shown in equations (38) and (39), (a / r Ｍ Values ​​other than those specified can be treated as constants determined by the reflection spectrum of the inspection light L1 and the characteristics of the dichroic mirror of the optical element 40. In other words, they can be considered known values. Therefore, the change in the relative values ​​of the light intensity ratio and light intensity difference is a value that depends on the material of the object S (a / r Ｍ It can be said that this is brought about solely by ).

[0070] In the inspection device 1A described above, the material of the foreign object W in the object S can be accurately identified by using characteristic data. Furthermore, the edge wavelength of the optical element 40 of the inspection device 1 is the edge wavelength (center wavelength λ) at which the light transmittance changes in a step-like manner from minimum to maximum, as shown in Figure 7. １ Rather than Δλ ａ (The smaller wavelength is λ) and the edge wavelength (center wavelength λ) is a wavelength greater than the edge wavelength, where the light transmittance changes in a step-like manner from maximum to minimum. １ Rather than Δλ ａIt includes ( / 2 larger wavelength). According to this, the step filter characteristics are center wavelength λ １ Range Δλ centered on the continent ａ This results in bandpass filter characteristics with a transmission region of . As a result, the degree of curvature of the incident light spectrum near the central wavelength of the inspection light can be reflected in the characteristic data, and the material of the foreign object can be identified based on the difference in the degree of curvature according to the material of the foreign object. Since this degree of curvature 'a' is a value that represents the material's properties themselves, it is effective in accurately performing quantitative analysis of materials. The target of quantitative analysis may be the composition of the material itself, or it may be an evaluation index for the functional aspects when using the material (for example, properties such as sugar, fat, water, protein, pH, or numerical values ​​of the freshness, sugar content, and acidity of food).

[0071] The optical element 40, instead of having multibandpass filter characteristics, has step filter characteristics in a predetermined wavelength range Δλ. ａ The filter may have multi-notch filter characteristics that include a cutoff region within which the transmittance of transmitted light Lt is minimized. In this case, the edge wavelengths corresponding to each center wavelength may include an edge wavelength (second edge wavelength) in which the transmittance of light changes in a step-like manner from the maximum value to the minimum value, and a third edge wavelength that is larger than the second edge wavelength and in which the transmittance of light changes in a step-like manner from the minimum value to the maximum value. [Third Embodiment]

[0072] The filter section 4 may be composed of three or more prisms. As shown in Figure 10, the inspection apparatus 1B according to the third embodiment differs from the inspection apparatus 1 in that the filter section 4A has an optical element 40A (first optical element), an optical element 40B (second optical element), a prism 41, a prism 42, and a prism 43, and the light detection section 5A has an image sensor 51, an image sensor 52, and an image sensor 53. The prisms 41, 42, and 43 are arranged to overlap each other in this order along the incident direction of the incident light L2.

[0073] Prism 41 has the shape of a roughly triangular prism with three sides: an input surface 41a perpendicular to the direction of incidence of incident light L2, a separation surface 41b inclined from the surface perpendicular to the direction of incidence of incident light L2, and an output surface 41c connecting the input surface 41a and the separation surface 41b. Optical element 40A is formed on the separation surface 41b. Prism 42 is optically coupled to prism 41 via the separation surface 41b. Prism 42 has the shape of a roughly triangular prism with three sides: an input surface 42a adjacent to the separation surface 41b, a separation surface 42b inclined from the surface perpendicular to the direction of incidence of incident light L2, and an output surface 42c connecting the input surface 42a and the separation surface 42b. Optical element 40B is formed on the separation surface 42b. Prism 43 is optically coupled to prism 42 via the separation surface 42b. The prism 43 has a roughly rectangular prism shape, including an input surface 43a adjacent to the separation surface 42b, and an output surface 43c facing the input surface 43a and substantially perpendicular to the incident direction of the incident light L2 as its side surface.

[0074] Optical elements 40A and 40B each have different step-type filter characteristics. Optical elements 40A and 40B are filters having step-type filter characteristics, such as dichroic mirrors. Dichroic mirrors are composed of, for example, dielectric multilayer films.

[0075] Figure 11(a) shows an example of the step filter characteristics of optical element 40A. Figure 11(b) shows an example of the step filter characteristics of optical element 40B. The step filter characteristics of optical elements 40A and 40B are centered at the wavelength λ １ ~λ ５ From each of these, within a predetermined wavelength range Δλ ｂ This exhibits multi-edge filter characteristics, showing edge filter properties at shifted wavelengths. The range here is Δλ. ｂ This refers to the full width at half maximum Δλ ｗ , and the range Δλ in the multibandpass filter shown in Figure 7. ａ Smaller than Δλ. ｂ is the range Δλ ａIt may be less than half of that. The edge wavelength of optical element 40A is different from the edge wavelength of optical element 40B (the fourth edge wavelength). For example, the center wavelength λ １ The edge wavelength of the optical element 40A corresponding to this is the wavelength (λ) as shown in Figure 11(a). １ -Δλ ｂ ) is the case. In contrast, the central wavelength λ １ The edge wavelength of the optical element 40B corresponding to this is the wavelength (λ) as shown in Figure 11(b). １ +Δλ ｂ ) is the case. In the example in Figure 11(a), the wavelength (λ １ -Δλ ｂ ) from wavelength (λ ２ +Δλ ｂ ) up to wavelength range, wavelength (λ ３ -Δλ ｂ ) from wavelength (λ ４ +Δλ ｂ ) up to the wavelength range, and wavelength (λ ５ -Δλ ｂ The wavelength region greater than λ is the transmission region, and the wavelength region other than these is the cutoff region. In contrast, in the example of Figure 11(b), the wavelength (λ １ +Δλ ｂ ) Wavelength region smaller than wavelength (λ ２ -Δλ ｂ ) from wavelength (λ ３ +Δλ ｂ ) up to the wavelength range, and wavelength (λ ４ -Δλ ｂ ) from wavelength (λ ５ +Δλ ｂ The wavelength range up to ) is the transmission region, and the wavelength range outside of this range is the cutoff region.

[0076] The reflected light Lr reflected in the filter section 4A includes reflected light Lr1 and reflected light Lr2. The transmitted light Lt that passes through the filter section 4A includes transmitted light Lt1 and transmitted light Lt2. The prisms 41, 42, and 43 of the filter section 4A function as follows: Incident light L2 is incident on prism 41 from the input surface 41a. The prism 41 has a transmission region (wavelength region where the transmittance is at its maximum) in the step-shaped filter characteristics of the optical element 40A of the incident light L2 that has passed through the input surface 41a and the inside of the medium of the prism 41, and generates transmitted light Lt1 that has passed through the optical element 40A. At the same time, the prism 41 has a cutoff region (wavelength region where the transmittance is at its minimum) in the optical element 40A, and generates reflected light Lr1 that has been reflected at the separation surface 41b. The reflected light Lr1 is reflected by the optical element 40A, then undergoes further total internal reflection at the input surface 41a, and is output to the outside from the output surface 41c.

[0077] In the prism 42, transmitted light Lt1 is incident from the input surface 42a. The prism 42 then has a transmission region in the step-shaped filter characteristics of the optical element 40B of the transmitted light Lt1 that has passed through the input surface 42a and the inside of the medium of the prism 42, and transmits transmitted light Lt2 that has passed through the separation surface 42b. At the same time, the prism 42 has a cutoff region in the step-shaped filter characteristics of the optical element 40B, and transmits reflected light Lr2 that has been reflected at the separation surface 42b. The reflected light Lr2 is reflected by the optical element 40B and then totally reflected at the input surface 42a and output to the outside from the output surface 42c. This total reflection of the reflected light Lr2 at the input surface 42a is achieved by forming an air gap between the prism 41 and the prism 42, or by applying a mirror finish to the reflection point of the reflected light Lr2 at the input surface 42a. In the prism 43, transmitted light Lt2 is received at the input surface 43a, and the transmitted light Lt2 that has passed through the input surface 43a and the inside of the medium of the prism 43 is output to the outside from the output surface 43c.

[0078] As shown in Figure 10, the light detection unit 5A includes an image sensor 51, an image sensor 52, and an image sensor 53. Image sensor 51 is positioned close to the output surface 41c of the prism 41, image sensor 52 is positioned close to the output surface 42c of the prism 42, and image sensor 53 is positioned close to the output surface 43c of the prism 43. Image sensor 53 is, for example, a line sensor, similar to image sensors 51 and 52. Image sensor 51 outputs luminance data based on the luminance of the received reflected light Lr1 to the data generation unit 6. Image sensor 52 outputs luminance data based on the luminance of the received reflected light Lr2 to the data generation unit 6. Image sensor 53 outputs luminance data based on the luminance of the received transmitted light Lt2 to the data generation unit 6. In the third embodiment, for example, if reflected light Lr1 is the first light, then reflected light Lr2 or transmitted light Lt2 becomes the second light.

[0079] Image sensor 51 detects reflected light Lr1. Reflected light Lr1 is light corresponding to the cutoff region of the step filter characteristics of optical element 40A. Image sensor 52 detects reflected light Lr2. Reflected light Lr2 is light corresponding to the cutoff region of the step filter characteristics of optical element 40B. Image sensor 53 detects transmitted light Lt2. Transmitted light Lt2 is light corresponding to the transmission region in the region where the step filter characteristics of optical element 40A and the step filter characteristics of optical element 40B overlap. That is, transmitted light Lt2 is light corresponding to the transmission region of the multibandpass filter characteristics as shown in Figure 7.

[0080] The data generation unit 6 in the inspection device 1B may combine the light quantities of each light as follows when generating light quantity difference image data GC by determining the light quantity of each light based on the brightness data output from the image sensors 51 to 53. The data generation unit 6 determines the light quantity of reflected light Lr1 as the light quantity of reflected light Lr I Ｒ The sum of the amount of reflected light Lr2 and the amount of transmitted light Lt2 is the amount of transmitted light Lt I. Ｔ This may also be done. The data generation unit 6 calculates the sum of the amount of reflected light Lr1 and the amount of transmitted light Lt2 as the amount of reflected light Lr I ＲThe amount of reflected light Lr2 is the amount of transmitted light Lt I Ｔ This may also be done. The data generation unit 6 calculates the sum of the light intensity of reflected light Lr1 and the light intensity of reflected light Lr2 as the light intensity of reflected light Lr I Ｒ Let the amount of transmitted light Lt2 be the amount of transmitted light Lt I Ｔ That is also acceptable.

[0081] Details of the light intensity ratio image data GL and light intensity difference image data GC in the data generation unit 6 of the inspection device 1B will be explained. Among the combinations of light intensity of each light described above, the central wavelength λ １ and central wavelength λ ３ Regarding this, the data generation unit 6 calculates the sum of the amount of reflected light Lr1 and the amount of transmitted light Lt2 as the amount of light I Ｒ Let the amount of reflected light Lr2 be the amount of light I. Ｔ Let's explain the case where this is the case. For the explanation, we will assume the step filter characteristics shown in Figure 12(a), the reflection spectrum of the inspection light L1 shown in Figure 12(b), and the spectral reflection spectrum of the object S shown in Figure 13. As shown in Figure 12(a), the transmittance of the filter section 4 is wavelength λ Ｍａ The wavelength increases from 0% to 100% as the edge wavelength λ. Ｍａ The central wavelength is λ Ｍ From the wavelength range Δλ ｂ This wavelength is shifted to the shorter wavelength side. Wavelength λ Ｍａ Smaller wavelengths than λ Ａ Then the transmittance is 0%, and the wavelength λ Ｍａ Larger wavelengths than λ Ｂ Therefore, the transmittance is 100%. The spectral reflectance spectrum shown in Figure 12(b) is the same as the spectral reflectance spectrum shown in Figure 5(b).

[0082] As shown in Figure 13, the spectral reflectance spectrum of the object S is, wavelength λ Ａ from wavelength λ Ｍａ The reflectivity gradually decreases as the wavelength λ approaches. Ｍａ from wavelength λ Ｂ It exhibits a nonlinear characteristic where the reflectance gradually increases over time. In the example in Figure 13, the approximate r ０ (λ) is a linear function approximation and k can be obtained by equations (19) and (20). ａ ,rＭ１ａ Using this, it is expressed by equation (43). ０ (λ) = k ａ (λ-λ) Ｍａ ) + r Ｍ１ａ ... (43) In equation (43), r Ｍ１ａ is the wavelength λ Ｍａ This is the reflectance at k. ａ is the wavelength λ Ｍａ This is the rate of change of the spectral reflectance spectrum with respect to wavelength in the vicinity, or in other words, wavelength λ Ｍａ This is the slope of the spectral reflectance spectrum in the vicinity. ａ This is expressed by equation (44). k ａ =Δr / Δλ...(44)

[0083] Based on equations (43) and (44), the light intensity ratio is (I Ｔ / I Ｒ ), and the relative value of the difference in light intensity (I Ｔ -I Ｒ ) / (2 × (I Ｔ +I Ｒ )) is calculated by the following formulas (45) and (46). As shown in equations (45) and (46), (k ａ / r Ｍ１ａ Other values ​​can be treated as constants determined by the reflection spectrum of the inspection light L1 and the characteristics of the dichroic mirror of the optical element 40. In other words, they can be considered known values. Therefore, the change in the relative values ​​of the light intensity ratio and light intensity difference is a value that depends on the material of the object S (k ａ / r Ｍ１ａ It can be said that this is brought about solely by the following. Of the combinations of light intensity for each light, the data generation unit 6 calculates the sum of the light intensity of reflected light Lr1 and the light intensity of transmitted light Lt2 as the light intensity of reflected light Lr I Ｒ The amount of reflected light Lr2 is the amount of transmitted light Lt I Ｔ Similarly, in this case, the relative values ​​of the light intensity ratio and light intensity difference are calculated using equations (45) and (46). Furthermore, among the combinations of light intensity for each light, the data generation unit 6 calculates the sum of the light intensity of reflected light Lr1 and the light intensity of reflected light Lr2 as the light intensity of reflected light Lr I Ｒ Let the amount of transmitted light Lt2 be the amount of transmitted light Lt I ＴIn this case, the transmitted light Lt2 becomes the light corresponding to the transmission region of the multibandpass filter characteristics. This approximates r ０ (λ) is approximated by a quadratic function as shown in Figure 9, and the relative values ​​of the light intensity ratio and light intensity difference are calculated by equations (38) and (39).

[0084] Of the combinations of light intensity for each of the above-mentioned types of light, the central wavelength λ １ and central wavelength λ ３ Regarding this, the data generation unit 6 calculates the amount of reflected light Lr1 as light quantity I Ｒ The sum of the amount of reflected light Lr2 and the amount of transmitted light Lt2 is the amount of light I. Ｔ In this case, as shown in Figure 14(a), the transmittance of the filter section 4 is wavelength λ Ｍｂ The value is increased from 0% to 100% with respect to the edge wavelength. Here, wavelength λ Ｍｂ The central wavelength is λ Ｍ From the wavelength range Δλ ｂ This wavelength is shifted to the longer wavelength side. In this case, the approximate r ０ (λ) is a linear function approximation, as shown in Figure 14(b), and k can be obtained by equations (19) and (20). ｂ ,r Ｍ１ｂ It is calculated using equation (47). Also, the light intensity ratio is (I Ｔ / I Ｒ ), and the relative value of the difference in light intensity (I Ｔ -I Ｒ ) / (2 × (I Ｔ +I Ｒ )) is calculated by the following formulas (48) and (49).

[0085] Of the combinations of light intensity for each of the above-mentioned types of light, the central wavelength λ １ and central wavelength λ ３ Regarding this, the data generation unit 6 calculates the sum of the light intensities of reflected light Lr1 and reflected light Lr2 as light intensity I Ｒ Let the amount of transmitted light Lt2 be the amount of light I Ｔ In this case, as shown in Figure 15(a), the characteristics of the filter section 4 are wavelength λ Ｍａ and wavelength λ Ｍｂ This results in a bandpass filter characteristic for the bandwidth between [value] and [value]. In this case, the approximate r ０(λ) is a quadratic function approximation, as shown in Figure 15(b), and a and r can be obtained by equations (40) and (41). Ｍ２ It is calculated using equation (50). Also, the light intensity ratio is (I Ｔ / I Ｒ ), and the relative value of the difference in light intensity (I Ｔ -I Ｒ ) / (2 × (I Ｔ +I Ｒ )) is calculated by the following formulas (51) and (52).

[0086] The approximation r of the linear function obtained from equations (43) and (47) ０ The approximation of the quadratic function r obtained from (λ) and equation (50) ０ By combining the results of both (λ) and (λ), it is possible to improve the accuracy of the approximation. Approximation r ０１２ (λ) is calculated by equation (53). Using equation (53) to improve the accuracy of the approximation is effective when quantitative analysis of materials requires properties such as sugar, fat, moisture, protein, and pH. The target of quantitative analysis may be the composition of the material itself, or it may be an evaluation index for the functional aspects of using the material (for example, properties such as sugar, fat, moisture, protein, and pH, or numerical values ​​of the freshness, sugar content, and acidity of food).

[0087] In the inspection device 1B described above, the material of the foreign object W in the object S can be accurately identified by using characteristic data. Furthermore, the filter unit 4A further includes an optical element 40B having step-shaped filter characteristics at an edge wavelength different from the edge wavelength of the optical element 40A. With this, multiple characteristic data can be generated by combining the luminance data of transmitted light Lt1 and reflected light Lr1 generated by the optical element 40A with the luminance data of transmitted light Lt2 and reflected light Lr2 generated by the optical element 40B, respectively, thereby enabling even more accurate identification of the material of the foreign object W. [Fourth Embodiment]

[0088] The light detection unit 5 may be configured as a multi-line sensor. In addition, the filter unit 4 does not have to be composed of overlapping prisms, but may instead be composed of, for example, a plate-type filter of parallel plane plates. In that case, the system can be configured by placing existing single-plate cameras on both the transmission side and the reflection side of the filter unit. By using a single-plate camera and a plate-type filter of parallel plane plates, the filter unit can be miniaturized, shortening the development period and reducing development costs. As shown in Figure 16, in the light detection unit 5B according to the fourth embodiment, optical elements 40C and 40D are formed on the light detection unit 5B. Optical element 40C is, for example, a dichroic mirror having the multi-edge filter characteristics shown in Figure 4. Optical element 40D is, for example, a dichroic mirror having the multi-bandpass filter characteristics shown in Figure 7. The light detection unit 5B includes an image sensor 54 that receives incident light L2 from an object S without passing through optical elements 40C and 40D using a plurality of pixels arranged in one direction, an image sensor 55 that receives transmitted light Lt that has passed through optical element 40C using a plurality of pixels arranged in the same direction as the image sensor 54, and an image sensor 56 that receives transmitted light Lt that has passed through optical element 40D using a plurality of pixels arranged in the same direction as the image sensors 54 and 55. For example, the light detection unit 5B may be an area sensor, and each image sensor 54, 55, and 56 may be a line of some pixels in the area sensor. In other words, each image sensor 54, 55, and 56 may be a line-shaped sensor portion composed of a plurality of pixels arranged in one direction in the area sensor. In this case, luminance data may be selectively read from a line of some pixels in the area sensor. The image sensors 54, 55, and 56 are arranged along directions that intersect in one direction, and constitute a multi-line sensor.

[0089] In the fourth embodiment, instead of the light detection unit 5 and filter unit 4 shown in Figure 1, a light detection unit 5B and optical elements 40C and 40D with the above configuration are provided. The data generation unit 6 subtracts the luminance of the transmitted light Lt received by the image sensor 55 from the luminance of the incident light L2 received by the image sensor 54, thereby indirectly determining the luminance of the reflected light Lr reflected by the optical element 40C. In the example in Figure 16, the transmitted light Lt that has passed through the optical element 40C of the incident light L2 is received by the image sensor 55 as the first light, or the transmitted light Lt that has passed through the optical element 40D of the incident light L2 is received by the image sensor 56 as the first light, and the incident light L2 is received by the image sensor 54 as the second light. The incident light L2 is assumed to include reflected light Lr. Since the image sensor 54 receives the incident light L2 without passing it through the optical elements, the amount of light detected by the image sensor 54 is greater than that detected by the image sensor 55 and the image sensor 56. Therefore, the exposure time of the image sensor 54 is set to be shorter than the exposure times of the image sensors 55 and 56, and sensitivity correction may be performed on the image sensor 54. Alternatively, instead of shortening the exposure time, sensitivity correction may be performed by arranging an ND filter with uniform transmittance across the entire effective wavelength range. The data generation unit 6 generates light intensity difference image data GC for inspecting the object S based on the light reception result at the light detection unit 5B. Since the optical element 40C has multi-edge filter characteristics, the light intensity difference image data GC reflects the slope component of the spectrum of the incident light L2 (spectral reflectance spectrum of the object S) near the central wavelength of the inspection light L1. Similarly, the light detection unit 5B can indirectly determine the brightness of the reflected light Lr reflected by the optical element 40D by subtracting the brightness of the transmitted light Lt received by the image sensor 56 from the brightness of the incident light L2 received by the image sensor 54. The data generation unit 6 generates light intensity difference image data GC for inspecting the object S based on the light reception results from the light detection unit 5B. Since the optical element 40D has multibandpass filter characteristics, the light intensity difference image data GC reflects the degree of curvature of the spectrum of the incident light L2 near the central wavelength of the inspection light L1.

[0090] The light detection unit 5B is configured with multiple image sensors, including an image sensor 55 that receives transmitted light Lt transmitted through the optical element 40C by multiple pixels arranged in one direction, arranged in a direction that intersects in one direction. This configuration allows the light detection unit 5B to be made compact by using image sensors 54, 55, and 56. Furthermore, luminance data corresponding to reflected light Lr can be obtained by the difference between the luminance data output from image sensor 54 and the luminance data of transmitted light Lt output from image sensor 55. Similarly, luminance data corresponding to reflected light Lr can be obtained by the difference between the luminance data output from image sensor 54 and the luminance data of transmitted light Lt output from image sensor 56. Additionally, as the object S is transported by the transport unit 2 in a direction that intersects in one direction, the object S can be scanned by the inspection device including the light detection unit 5B. [Fifth Embodiment]

[0091] As shown in Figure 17(a), the photodetector 5C according to the fifth embodiment functions as a multi-line sensor, and optical elements 40E and 40F are formed on the photodetector 5C. The optical element 40E has a center wavelength λ, as shown in Figure 17(b), for example. １ and central wavelength λ ５ At this point, the transmittance increases from 0% to 100%, and the central wavelength λ ４ This is a dichroic mirror having multi-edge filter characteristics that exhibit a characteristic where the transmittance decreases from 100% to 0% at a certain point. The optical element 40F has a center wavelength λ, as shown in Figure 17(c), for example. ３ At this point, the transmittance increases from 0% to 100%, and the central wavelength λ ２This is a dichroic mirror having multi-edge filter characteristics that exhibit a characteristic where the transmittance decreases from 100% to 0%. The light detection unit 5C includes an image sensor 55 that receives transmitted light Lt that has passed through the optical element 40E by a plurality of pixels arranged in one direction, and an image sensor 56 that receives transmitted light Lt that has passed through the optical element 40F by a plurality of pixels arranged in the same direction as the image sensor 55. For example, similar to the light detection unit 5B, the light detection unit 5C may be an area sensor, and each image sensor 55, 56 may be a line of some pixels in the area sensor.

[0092] In the fifth embodiment, instead of the light detection unit 5 and filter unit 4 shown in Figure 1, a light detection unit 5C and optical elements 40E and 40F with the above configuration are provided. In this case, the data generation unit 6 subtracts the luminance of the transmitted light Lt received by the image sensor 56 from the luminance of the transmitted light Lt received by the image sensor 55, thereby determining the center wavelength λ ２ From the central wavelength λ ３ The brightness of light transmitted through the wavelength range between these two wavelengths can be indirectly determined as the brightness of the reflected light Lr. Furthermore, the data generation unit 6 subtracts the brightness of the transmitted light Lt received by the image sensor 55 from the brightness of the transmitted light Lt received by the image sensor 56, thereby determining the central wavelength λ. １ A wavelength range smaller than λ, and a central wavelength λ ４ From the central wavelength λ ５ The brightness of light transmitted through the wavelength range between these wavelengths can be indirectly determined as the brightness of reflected light Lr. The data generation unit 6 generates light intensity difference image data GC and light intensity ratio image data GL for inspection of the object S based on the light reception results from the light detection unit 5C.

[0093] As described above, even when the light detection unit is a multi-line sensor, such as the light detection unit 5B in the fourth embodiment and the light detection unit 5C in the fifth embodiment, the material of the foreign object W in the target object S can be accurately identified by using characteristic data. [Sixth Embodiment]

[0094] The inspection apparatus 1C according to the sixth embodiment differs from the inspection apparatus 1 according to the first embodiment in that, as shown in Figure 18, it further comprises a polarizing plate 91 positioned before the filter section 4 in the optical path of the incident light L2, and a polarization rotation section 92 positioned between the filter section 4 and the polarizing plate 91. In the inspection apparatus 1C, the data generation section 6 generates different light intensity difference image data GC according to the polarization direction of the incident light L2. The polarization state of the incident light L2 is linearly polarized, and the incident light L2 includes a p-polarization component and an s-polarization component. The p-polarization component is the polarization component in which the electric field vector in the incident light L2 vibrates parallel to the incident plane. The s-polarization component is the polarization component in which the electric field vector in the incident light L2 vibrates perpendicular to the incident plane. Here, the incident plane is the plane formed by the direction of propagation of the incident light L2 and the thickness direction (depth direction) of the filter section 4. The polarization direction of the p-polarization component is orthogonal to the polarization component of the s-polarization component.

[0095] The polarizing plate 91 transmits either the p-polarized component or the s-polarized component of the incident light L2. For example, if the transmission axis of the polarizing plate 91 is perpendicular to the incident plane, it transmits s-polarized light, and if the transmission axis of the polarizing plate 91 is parallel to the incident plane, it transmits p-polarized light. The polarization rotation unit 92 rotates the polarization direction of the polarization components that have passed through the polarizing plate 91. The polarization rotation unit 92 may output the polarization component of the incident light L2 as a p-polarized component to the filter unit 4 by rotating the polarization direction of the s-polarized component by 90°. Alternatively, the polarization rotation unit 92 may output the polarization component of the incident light L2 as an s-polarized component to the filter unit 4 by rotating the polarization direction of the p-polarized component by 90°. The polarization rotation unit 92 may be a half-wave plate or a polarization rotator.

[0096] The optical element 40 included in the filter section 4 has different step-like filter characteristics depending on the polarization direction of the incident light L2, for example. As shown in Figure 19(a), the optical element 40 exhibits filter characteristic FR1 for the p-polarized component and filter characteristic FR2 for the s-polarized component. In filter characteristic FR1, the wavelength (λ) １ -Δλ ｂ ) from wavelength (λ ２ +Δλ ｂ ) up to wavelength range, wavelength (λ ３-Δλ ｂ The wavelength region greater than λ is the transmission region, and the wavelength region other than these is the cutoff region. In contrast, in the filter characteristic FR2, the wavelength (λ) １ +Δλ ｂ ) from wavelength (λ ２ -Δλ ｂ ) up to wavelength range, wavelength (λ ３ +Δλ ｂ The wavelength region greater than ) is the transmission region, and the wavelength region other than these is the cutoff region. Figure 19(b) shows an example of the spectral reflectance spectrum of the object S as a function r(λ). Comparing Figure 19(a) and Figure 19(b), at the central wavelength λ1, the approximate r of the function r(λ) is ０ (λ) is a linear approximation, whereas the central wavelength λ ２ and λ ３ So, the approximation of the function r(λ) is r ０ (λ) is a quadratic function approximation.

[0097] In the inspection device 1C, the data generation unit 6 may combine the light intensities of each light as follows when generating light intensity difference image data GC based on different step-shaped filter characteristics according to the polarization direction of the incident light L2. The data generation unit 6 determines the light intensity of the reflected light Lrp of the p-polarization component and the light intensity of the reflected light Lr I Ｒ Let the amount of transmitted light Ltp of the p-polarization component be the amount of transmitted light Lt I Ｔ Alternatively, the data generation unit 6 may calculate the amount of reflected light Lrs of the s-polarization component as the amount of reflected light Lr I Ｒ Let the amount of transmitted light Lts of the s-polarization component be the amount of transmitted light Lt I Ｔ This may also be done. In the above combination, filter characteristics FR1 and FR2 are considered to be multi-edge filter characteristics, and the data generation unit 6 approximates the function r(λ) as shown in equation (1). ０ The light intensity difference image data GC may be generated assuming that (λ) is a linear approximation. Alternatively, the data generation unit 6 may calculate the light intensity I as the sum of the light intensity of the reflected light Lrp of the p-polarization component and the light intensity of the transmitted light Lts of the s-polarization component. Ｒ The difference between the amount of transmitted light Ltp of the p-polarized component and the amount of transmitted light Lts of the s-polarized component is the amount of transmitted light Lt I. ＴThis may also be done. In the above combination, the filter characteristics FR1 and FR2 are combined and considered as a multibandpass filter characteristic, and the data generation unit 6 approximates the function r(λ) as shown in equation (22). ０ The light intensity difference image data GC may be generated by assuming that (λ) is a quadratic function approximation.

[0098] In the inspection device 1C described above, the material of the foreign object W in the object S can be accurately identified by using characteristic data. Furthermore, the inspection device 1C is further equipped with a polarization rotation unit 92 positioned before the filter unit 4 in the optical path of the incident light L2, which rotates the polarization direction of the incident light L2. As a result, the filter unit 4 exhibits different filter characteristics depending on the polarization direction of the incident light L2. This allows for the generation of multiple characteristic data, enabling even more accurate identification of the material of the foreign object W. [Inspection Method]

[0099] Figure 20 is a flowchart showing an inspection method according to one embodiment of the present disclosure. As shown in Figure 20, the inspection method according to this embodiment comprises an output step S01, a light guiding step S02, a light detection step S03, a generation step S04, an identification step S05, and a display step S06. The inspection method is carried out, for example, using the inspection devices 1, 1A, 1B, and 1C described above.

[0100] Output step S01 is a step in which inspection light L1 is output toward the object S by switching wavelengths over time. In output step S01, for example, each LED bar of the light source unit 3, which is composed of a multiband LED bar illumination device, is lit sequentially, and inspection light L1 is output toward the object S by switching wavelengths over time.

[0101] The light guiding step S02 is a step in which incident light L2 from the object S is guided to an optical element 40 having a step-like filter characteristic in which the transmittance of light changes in a step-like manner between a maximum value and a minimum value in a predetermined wavelength region including the central wavelength of the inspection light L1. In the light guiding step S02, depending on the transmission characteristics of the optical element 40, a portion of the incident light L2 from the object S becomes transmitted light Lt, and the remainder of the incident light L2 from the object S becomes reflected light Lr.

[0102] The light detection step S03 is a step in which the transmitted light Lt that has passed through the optical element 40 and the reflected light Lr that has been reflected by the optical element 40 from the incident light L2 from the object S are received, and the brightness data of the transmitted light Lt and the brightness data of the reflected light Lr are output. For example, in the inspection device 1, the transmitted light Lt that has passed through the optical element 40 from the incident light L2 from the object S is received by the image sensor 52, and the brightness data of the transmitted light Lt is output to the computer 10. In addition, the reflected light Lr that has been reflected by the optical element 40 from the incident light L2 from the object S is received by the image sensor 51, and the brightness data of the reflected light Lr is output to the data generation unit 6 of the computer 10.

[0103] The generation step S04 is a step in which light intensity difference image data GC of the object S with respect to incident light L2 from the object S is generated based on the luminance data of transmitted light Lt and the luminance data of reflected light Lr. In the generation step S04, the data generation unit 6 of the computer 10 generates the light intensity difference image data GC and outputs the generated light intensity difference image data GC to the identification unit 7.

[0104] Identification step S05 is a step in which the material of the foreign object W in the target object S is identified based on the light intensity difference image data GC. In identification step S05, the identification unit 7 of the computer 10 analyzes the light intensity difference image data GC in each wavelength region and identifies the material of the foreign object W in the target object S. Then, information indicating the identification result is output to the display unit 8.

[0105] Display step S06 is a step in which the identification result of the material of the foreign object W in the object S is displayed. In display step S06, the identification result output from the identification unit 7 is output to the display unit 8. In addition to displaying the identification result of the material of the foreign object W, in display step S06, light intensity ratio image data GL or light intensity difference image data GC may also be displayed together with the identification result.

[0106] Figure 21 is a flowchart showing an example of the details of the generation step S04 and identification step S05 described above. As shown in Figure 21, first, multiple luminance data for each wavelength region are acquired by sequential lighting of the light source unit 3 (step S11). Next, the acquired luminance data are combined for each wavelength region to obtain luminance data for each wavelength region (step S12). Furthermore, calculation processing is performed on the luminance data obtained by the combination to generate light intensity ratio image data GL or light intensity difference image data GC for each wavelength region (step S13). Then, based on the light intensity difference image data GC or light intensity ratio image data GL for each wavelength region, the material of the foreign object W in the target object S is identified (step S14). [First Modified Example]

[0107] The identification unit 7 may determine whether the material of the foreign object W in the object S is transparent based on a comparison between the light intensity ratio image data GL or the light intensity difference image data GC and a preset threshold. When the foreign object W is transparent, when the inspection light L1 is incident on the foreign object W, specular reflected light is generated that is specularly reflected at the surface of the foreign object W, specular reflected light is generated that is transmitted through the foreign object W and specularly reflected at the back surface, and multiple reflected light is generated when the light transmitted through the foreign object W is multiplely reflected within the foreign object W. These reflected lights reinforce or cancel each other out according to the phase difference between them which is determined by the wavelength and the optical path length, so when the detected incident light L2 is spectrally analyzed, a spectrum with periodic peaks and valleys on the wavelength axis is obtained. For example, as shown in Figure 22(a), when the light intensity ratio image data GL or the light intensity difference image data GC includes a transparent object Pt, as shown in Figures 22(b) and 22(c), a certain range (for example, the wavelength λ of the reflection spectrum of the incident light L2 shown in Figure 5) is obtained. Ａ from wavelength λ Ｂ Spectral reflectance R of transparent material Pt for wavelengths (a range greater than the wavelength range up to R) Ｐｔ A periodicity tends to appear. Here, the light incident on the filter section 4 has a spectral reflectance R Ｐｔ and the reflection spectrum S of the inspection light L1 Ｌ１It is expressed as the product of (represented by the dotted lines in Figures 22(b) and 22(c)). As described above, the light intensity ratio image data GL or the light intensity difference image data GC is data that shows the relative value between the luminance of transmitted light Lt and the luminance of reflected light Lr. For this reason, in the case of a transparent material Pt, if the filter characteristics of the filter section 4 are step-type filter characteristics F1, as shown in Figure 22(b), the luminance of transmitted light Lt that has passed through the filter section 4 may be significantly larger than the luminance of reflected light Lr that has been reflected by the filter section 4. Alternatively, if the filter characteristics of the filter section 4 are step-type filter characteristics F1, as shown in Figure 22(c), the luminance of transmitted light Lt that has passed through the filter section 4 may be significantly smaller than the luminance of reflected light Lr that has been reflected by the filter section 4.

[0108] Similarly, the filter characteristics of the filter section 4 are bandpass filter characteristics F2 (bandwidth is spectral reflectance R). Ｐｔ Even when the period is about half or less than half of the period, it is possible to determine whether the material of the foreign object W in the object S is transparent or not. In this case, with a transparent object Pt, as shown in Figure 23(a), the brightness of the transmitted light Lt that has passed through the filter part 4 may be significantly smaller relative to the brightness of the reflected light Lr that has been reflected by the filter part 4. Alternatively, as shown in Figure 23(b), the brightness of the transmitted light Lt that has passed through the filter part 4 may be significantly larger relative to the brightness of the reflected light Lr that has been reflected by the filter part 4.

[0109] Furthermore, in a multi-line sensor as shown in Figure 16, even when optical elements 40C and 40D are bandpass filters with different bandwidths, it is possible to determine whether the material of the foreign object W in the object S is transparent or not. As an example, as shown in Figure 24(a), the case where the filter characteristic of optical element 40C is a bandpass filter characteristic F3 and the filter characteristic of optical element 40D is a bandpass filter characteristic F4. The bandwidth of bandpass filter characteristic F3 is about half or less of the period of the spectral reflectance of the transparent material Pt, and the bandwidth of bandpass filter characteristic F4 is several times or more of the period of the spectral reflectance of the transparent material Pt. That is, the bandwidth of bandpass filter characteristic F3 is included in the bandwidth of bandpass filter characteristic F4. In this case, for the transparent material Pt, as shown in Figure 24(a), the brightness of the transmitted light Lt that has passed through optical element 40C may be significantly smaller relative to the brightness of the transmitted light Lt that has passed through optical element 40D. Alternatively, in the case of a transparent material Pt, as shown in Figure 24(b), the luminance of the transmitted light Lt that has passed through the optical element 40C may be significantly larger relative to the luminance of the transmitted light Lt that has passed through the optical element 40D. In this example, if the spectral width of the inspection light L1 is equivalent to the bandwidth of the bandpass filter characteristic F4, the luminance of the transmitted light Lt that has passed through the optical element 40D, i.e., the output of the image sensor 56, may be replaced with the luminance detected without passing through the optical element, i.e., the output of the image sensor 54.

[0110] When considering such trends, the identification unit 7 sets an upper threshold X and a lower threshold Y for the luminance value of, for example, the light intensity ratio image data GL or the light intensity difference image data GC. The identification unit 7 determines that the foreign object W is a non-transparent material Pn if the luminance value of the light intensity ratio image data GL or the light intensity difference image data GC is between Y and X, and determines that the foreign object W is a transparent material Pt if the luminance value of the light intensity ratio image data GL or the light intensity difference image data GC is less than Y or greater than X. In this way, by applying thresholds to the light intensity ratio image data GL or the light intensity difference image data GC and detecting the bias between the luminance of transmitted light Lt and the luminance of reflected light Lr, it is possible to determine with high accuracy whether the material of the foreign object W in the target object S is a transparent material Pt or not.

[0111] Figure 25 is a flowchart showing an example of an inspection method according to the first modified example. Steps S31 to S33 shown in Figure 25 are the same as steps S11 to S13 shown in Figure 21, so their explanation is omitted. After generating light intensity ratio image data GL and light intensity difference image data GC for each wavelength region, the brightness value of the generated light intensity difference image data GC is compared with a preset threshold (here, an upper threshold and a lower threshold) (step S34), and it is determined whether or not the brightness value of the light intensity difference image data GC is within the threshold range (step S35). In step S35, if the brightness value of the light intensity difference image data GC is within the threshold range, it is determined that the foreign object W is an opaque object (step S36). In step S35, if the brightness value of the light intensity difference image data GC is not within the threshold range, it is determined that the foreign object W is a transparent object Pt (step S37). Furthermore, when determining whether or not the foreign object W is a transparent material Pt, it is also possible to first determine whether or not the foreign object W is a transparent material Pt, and then perform the material determination based on the correlation described above for foreign object W that is determined to be opaque. This allows for efficient identification of the material of the foreign object W. [Second Modified Example]

[0112] When the foreign object W contains a transparent material Pt, the intensity of the specularly reflected light tends to increase. Therefore, the inspection device may use the specularly reflected light of the transparent material Pt to determine whether or not it is a transparent material Pt. As shown in Figure 26, the inspection device 1D according to the second modified example differs from the inspection device 1 in that it further includes a polarizing plate 93 positioned after the light source devices 31 and 32 in the optical path of the inspection light L1, and that the filter section 4B is composed of prisms 41, 42, and 43. Furthermore, as shown in Figure 27, the filter section 4B differs from the filter section 4A according to the third embodiment in that a polarizing beam splitter BS is formed as an optical element 40A on the separation surface 41b. In addition, the step-shaped filter characteristics of the optical element 40B formed on the separation surface 42b of the filter section 4B are as shown in Figure 4, with a center wavelength λ １ ~λ ５Each of these may exhibit the characteristics of an edge filter, or it may be a multi-edge filter characteristic as shown in Figure 7, or a multi-bandpass filter characteristic.

[0113] The polarizing beam splitter BS reflects the polarization component of the incident light L2 that is parallel to the transmission axis of the polarizer 93, and transmits the polarization component of the incident light L2 that is perpendicular to the transmission axis of the polarizer 93. In the inspection device 1D, the polarizing beam splitter BS and the polarizer 93 function as orthogonal nicols or parallel nicols with respect to the incident light L2. In the inspection device 1D, by having the polarizing beam splitter BS and the polarizer 93 function as orthogonal nicols with respect to the incident light L2, specular reflected light is blocked, and only scattered light, as described later, is detected. The data generation unit 6 generates light intensity difference image data GC based only on the spectrum of scattered light, and the identification unit 7 identifies the material of the foreign object W in the object S based on the light intensity difference image data GC, thereby improving the accuracy of the determination. On the other hand, in the inspection device 1D, by having the polarizing beam splitter BS and the polarizer 93 function as parallel nicols with respect to the incident light L2, specular reflected light can be detected efficiently.

[0114] Figure 28 is a diagram for conceptually explaining orthogonal nicols. In the example in Figure 28, incident light L2 that has passed through the polarizer 93 and been reflected by the object S is shown as incident on the polarizer 93A. The transmission axis of the polarizer 93 is perpendicular to the transmission axis of the polarizer 93A. Figure 28(a) is a diagram to explain the case in which the incident light L2 is specularly reflected by the object S, and Figure 28(b) is a diagram to explain the case in which the incident light L2 is scattered by the object S. As shown in Figure 28(a), the polarization direction of the inspection light L1 output from the light source device 31 is randomly diverse. When the inspection light L1 is incident on the polarizer 93 in this state, the inspection light L1 is output from the polarizer 93 in a polarization direction D1 parallel to the transmission axis of the polarizer 93. When the inspection light L1 is specularly reflected on the surface of the object S, the polarization direction of the specularly reflected incident light L2 is parallel to the polarization direction D1 of the inspection light L1 that has passed through the polarizer 93. However, since the transmission axis of the polarizer 93A is perpendicular to the polarization direction D1, the specularly reflected incident light L2 is reflected or absorbed by the polarizer 93A. Alternatively, the specularly reflected incident light L2 is blocked by the polarizer 93A. In contrast, as shown in Figure 28(b), when the inspection light L1 output from the polarizer 93 in polarization direction D1 is scattered inside the object S, the scattered incident light L2 is transmitted through the polarizer 93A in a polarization direction parallel to the transmission axis of the polarizer 93A, that is, in a polarization direction D2 perpendicular to polarization direction D1.

[0115] Figure 29 is a diagram for conceptually explaining parallel nicols. In the example in Figure 29, incident light L2 that has passed through the polarizer 93 and been reflected at the object S is shown as incident light on the polarizer 93B. The transmission axis of the polarizer 93 is parallel to the transmission axis of the polarizer 93B. Figure 29(a) is a diagram for explaining the case where the inspection light L1 is specularly reflected at the object S, and Figure 29(b) is a diagram for explaining the case where the inspection light L1 is scattered at the object S. As shown in Figure 29(a), when the inspection light L1 output from the polarizer 93 in polarization direction D1 is specularly reflected at the surface of the object S, the incident light L2, which is specularly reflected light, passes through the polarizer 93B in polarization direction D1 parallel to the transmission axis of the polarizer 93B. In contrast, as shown in Figure 29(b), when the inspection light L1 output from the polarizer 93 in polarization direction D1 is scattered inside the object S, the scattered incident light L2 is transmitted through the polarizer 93B in polarization direction D1 parallel to the transmission axis of the polarizer 93B. Whether the incident light L2 is specularly reflected or scattered, it is transmitted through the polarizer 93B. However, since the intensity of the scattered light is smaller than the intensity of the specularly reflected light, the incident light L2 transmitted through the polarizer 93B is predominantly specularly reflected.

[0116] The polarizing beam splitter BS shown in Figure 27 reflects the polarization component of the incident light L2 that is parallel to the transmission axis of the polarizer 93, and transmits the polarization component of the incident light L2 that is perpendicular to the transmission axis of the polarizer 93. Therefore, the light transmitted through the polarizer 93A in Figure 28 corresponds to the transmitted light Lt1 in Figure 27. The light transmitted through the polarizer 93B in Figure 29 corresponds to the reflected light Lr1 in Figure 27. The image sensor 51 detects the polarization component of the incident light L2 that is parallel to the transmission axis of the polarizer 93 as reflected light Lr1. The image sensor 52 detects the reflected light Lr2 from the transmitted light Lt1, which is the polarization component perpendicular to the transmission axis of the polarizer 93. The reflected light Lr2 is the light corresponding to the cutoff region in the step-shaped filter characteristics of the optical element 40B. The image sensor 53 detects the transmitted light Lt2 from the transmitted light Lt1. The transmitted light Lt2 is the light corresponding to the transmission region in the step-shaped filter characteristics of the optical element 40B. As mentioned above, when conceptually explaining parallel nicols or orthogonal nicols, polarizers 93A or 93B were used. However, in the inspection device 1D, polarizers 93A or 93B may be used instead of the polarizing beam splitter BS.

[0117] Figure 30 is a flowchart showing an example of an inspection method when determining whether or not a transparent material Pt is a transparent material Pt using the specular reflected light of the transparent material Pt with the inspection device 1D. Note that the identification method shown in Figure 30 may be performed instead of the identification method shown in Figure 25, or may be performed together with the identification method shown in Figure 25. Steps S41 and S42 shown in Figure 30 are the same as steps S11 and S12 shown in Figure 21, so their explanation is omitted. After obtaining luminance data for each wavelength region in step S42, calculation processing is performed on the luminance data obtained by synthesis to generate specular reflected image data GR for each wavelength region (step S43). The pixel value of each pixel included in the specular reflected image data GR is (I Ｒ１ / (I Ｒ２ +I Ｔ２ )) is expressed as the luminance value of the reflected light Lr1 received by each pixel of the image sensor 51, that is, the amount of reflected light Lr1 I Ｒ１The brightness value of the reflected light Lr2 received by each pixel of the image sensor 52, i.e., the amount of reflected light Lr2, is I Ｒ２ The brightness value of the transmitted light Lt2 received by each pixel of the image sensor 53, i.e., the amount of light of the transmitted light Lt2, is I Ｔ２ Next, the brightness value of the generated specular reflection image data GR is compared with a pre-set threshold (here, an upper threshold and a lower threshold) (step S44) to determine whether the brightness value of the specular reflection image data GR is equal to or greater than the threshold (step S45). In step S45, if the brightness value of the specular reflection image data GR is equal to or greater than the threshold, it is determined that the foreign object W is a transparent material Pt (step S46). In step S45, if the brightness value of the specular reflection image data GR is not equal to or greater than the threshold, it is determined that the foreign object W is an opaque material Pn (step S47).

[0118] In the inspection device 1D described above, the polarizing beam splitter BS may reflect one of the polarization components of the incident light L2 that is parallel to the transmission axis of the polarizing plate 93 and the polarization component that is perpendicular to the transmission axis of the polarizing plate 93, and transmit the other of the polarization components of the incident light L2 that is parallel to the transmission axis of the polarizing plate 93 and the polarization component that is perpendicular to the transmission axis of the polarizing plate 93. In this modified example, the case in which the polarizing beam splitter BS reflects the polarization component of the incident light L2 that is parallel to the transmission axis of the polarizing plate 93 and transmits the polarization component that is perpendicular to the transmission axis of the polarizing plate 93 has been described, but the polarizing beam splitter BS may also reflect the polarization component that is perpendicular to the transmission axis of the polarizing plate 93 and transmit the polarization component that is parallel to the transmission axis of the polarizing plate 93. In this case, the image sensor 51 detects the polarization component of the incident light L2 that is perpendicular to the transmission axis of the polarizing plate 93 as reflected light Lr1. Image sensor 52 detects reflected light Lr2 contained in transmitted light Lt1, which is the polarization component of the incident light L2 parallel to the transmission axis of the polarizer 93. Image sensor 53 detects transmitted light Lt2 contained in transmitted light Lt1.

[0119] In the inspection device 1D described above, the material of the foreign object W in the object S can be accurately identified by using characteristic data. Furthermore, the inspection device 1D further includes a polarizing plate 93 positioned after the light source devices 31 and 32 in the optical path of the inspection light L1, and the optical element 40A is a polarizing beam splitter BS that reflects one of the polarization components of the incident light L2 that is parallel to the transmission axis of the polarizing plate 93 and the polarization component that is perpendicular to the transmission axis of the polarizing plate 93, and transmits the other of the polarization components of the incident light L2 that is parallel to the transmission axis of the polarizing plate 93 and the polarization component that is perpendicular to the transmission axis of the polarizing plate 93. With this configuration, the polarizing plate 93 and the polarizing beam splitter BS have a parallel nicol configuration with respect to the polarization component parallel to the transmission axis of the polarizing plate 93, and a cross-nicol configuration with respect to the polarization component that is perpendicular to the transmission axis of the polarizing plate 93. As a result, in parallel nicols, which detect the polarization component parallel to the transmission axis of the polarizing plate 93, specular reflected light from the transparent material Pt can be efficiently detected, and it is possible to accurately determine whether or not the foreign material W contains the transparent material Pt. On the other hand, in orthogonal nicols, which detect the polarization component perpendicular to the transmission axis of the polarizing plate 93, the specular reflected component can be removed, and the accuracy of identifying foreign materials W other than the transparent material Pt can be improved.

[0120] Furthermore, in the inspection device 1D, the data generation unit 6 generates specular reflection image data GR based on the polarization component parallel to the transmission axis of the polarizing plate 93, and the identification unit 7 determines whether the object S contains transparent material Pt based on a comparison between the brightness of the specular reflection image data GR and a threshold. According to this, by making a determination based on the brightness of the specular reflection light, it is possible to determine with high accuracy whether the object S contains transparent material Pt. [Third Modified Example]

[0121] In the inspection device 1D, the light detection unit may be configured as a multi-line sensor. In addition, the filter unit does not have to be configured by overlapping prisms, but may instead be configured by, for example, a filter of parallel planar plates. As shown in Figure 31, the light detection unit 5D according to the third modified example is a multi-line sensor, and an optical element 40G, a polarizer 94, and a polarizer 95 are formed on the light detection unit 5D. The optical element 40G is, for example, a dichroic mirror having the multi-edge filter characteristics shown in Figure 4, or a dichroic mirror having the multi-bandpass filter characteristics shown in Figure 7. The transmission axis of the polarizer 94 is parallel to the transmission axis of the polarizer 93. That is, the polarizer 94 and the polarizer 93 function as parallel nicols with respect to the incident light L2. The transmission axis of the polarizer 95 is perpendicular to the transmission axis of the polarizer 93. That is, the polarizer 95 and the polarizer 93 function as orthogonal nicols with respect to the incident light L2. The light detection unit 5D includes an image sensor 57 that receives light from an object S that has passed through the polarizing plate 95, without going through the optical element 40G, using a plurality of pixels arranged in one direction. The light received by the image sensor 57 corresponds to the transmitted light Lt1 in Figure 27. The light detection unit 5D further includes an image sensor 58 that receives light from an incident light L2 that has passed through the polarizing plate 95 and also passed through the optical element 40G, using a plurality of pixels arranged in the same direction as the image sensor 57. The light received by the image sensor 58 corresponds to the transmitted light Lt2 in Figure 27. The light detection unit 5D includes an image sensor 59 that receives light from an incident light L2 that has passed through the polarizing plate 94, without going through the optical element 40G, using a plurality of pixels arranged in the same direction as the image sensors 57 and 58. The light received by the image sensor 59 corresponds to the reflected light Lr1 in Figure 27. For example, the light detection unit 5D may be an area sensor, and each image sensor 57, 58, 59 may be a line of pixels within the area sensor.

[0122] The data generation unit 6 may generate specular reflection image data GR based on the brightness of the reflected light Lr1 received by the image sensor 59 and the brightness of the transmitted light Lt1 received by the image sensor 57. Alternatively, the data generation unit 6 can indirectly determine the brightness of the reflected light Lr2 in Figure 27 by subtracting the brightness of the transmitted light Lt2 received by the image sensor 58 from the brightness of the transmitted light Lt1 received by the image sensor 57. Based on the light reception results of the light detection unit 5D, the data generation unit 6 generates light intensity difference image data GC for inspection of the object S. If the optical element 40G has multi-edge filter characteristics, the light intensity difference image data GC reflects the slope component of the spectrum of the incident light L2 near the central wavelength of the inspection light L1. If the optical element 40G has multi-bandpass filter characteristics, the light intensity difference image data GC reflects the degree of curvature of the spectrum of the incident light L2 near the central wavelength of the inspection light L1. [Fourth Modification]

[0123] The configuration for functioning as parallel nicols or orthogonal nicols with respect to incident light L2 is not limited to the inspection device 1D according to the second modification. As shown in Figure 32, the inspection device 1E according to the fourth modification differs from the inspection device 1D in that it includes a polarizing plate 96 positioned before the lens LZ2 in the optical path of the incident light L2, and that it includes a filter unit 4 instead of a filter unit 4B. The polarizing plate 96 is rotatable, for example, by a rotation drive unit (not shown). When the transmission axis of the polarizing plate 96 is parallel to the transmission axis of the polarizing plate 93, the polarizing plate 96 and the polarizing plate 93 function as parallel nicols with respect to incident light L2. As a result, when the incident light L2 is specularly reflected light, the incident light L2 passes through the polarizing plate 96 and enters the filter unit 4 via the lens LZ2. When the transmission axis of the polarizing plate 96 is orthogonal to the transmission axis of the polarizing plate 93, the polarizing plate 96 and the polarizing plate 93 function as orthogonal nicols with respect to incident light L2. As a result, when the incident light L2 is scattered light and the polarization direction of the scattered light is perpendicular to the transmission axis of the polarizer 93, the incident light L2 passes through the polarizer 96 and enters the filter section 4 via the lens LZ2. The data generation unit 6 may generate specular reflection image data GR based on the light reception result at the light detection unit 5. Note that the optical elements and light detection unit in the inspection device 1E are not limited to the filter section 4 and the light detection unit 5, but may be the filter section 4A and the light detection unit 5A as shown in Figure 10, or the optical elements 40C, 40D and the light detection unit 5B as shown in Figure 16. [Fifth Modification]

[0124] As shown in Figure 33, the inspection apparatus 1F according to the fifth modified example differs from the inspection apparatus 1E in that it is equipped with a fixed polarizing plate 97 instead of a rotating polarizing plate 96, and with a light source unit 3A instead of a light source unit 3. The light source unit 3A includes light source devices 31A, 31B, 32A, and 32B that can output inspection light L1 from four directions, upstream and downstream in the transport direction toward the object S. Between the light source device 31A and the object S, a polarizing plate 93A and a lens LZ1 are provided in this order in the direction of travel of the inspection light L1 emitted from the light source device 31A. Between the light source device 32A and the object S, a polarizing plate 93A and a lens LZ1 are provided in this order in the direction of travel of the inspection light L1 output from the light source device 32A. The transmission axis of the polarizing plate 93A is perpendicular to the transmission axis of the polarizing plate 97. That is, the polarizing plates 93A and 97 function as orthogonal nicols with respect to the incident light L2. Between the light source device 31B and the object S, a polarizing plate 93B and a lens LZ1 are provided in this order in the direction of travel of the inspection light L1 output from the light source device 31B. Between the light source device 32B and the object S, a polarizing plate 93B and a lens LZ1 are provided in this order in the direction of travel of the inspection light L1 emitted from the light source device 32B. The transmission axis of the polarizing plate 93B is parallel to the transmission axis of the polarizing plate 97. That is, the polarizing plates 93B and 97 function as parallel nicols with respect to the incident light L2. In the example of Figure 33, the light source devices 31A and 32A are provided such that the angle between the direction of travel of the inspection light L1 output from the light source devices 31A and 32A and the direction of travel of the incident light L2 is 45°. In contrast, the light source devices 31B and 32B are provided such that the angle between the direction of travel of the inspection light L1 output from the light source devices 31B and 32B and the direction of travel of the incident light L2 is less than 45°.

[0125] The light source unit 3A, like the light source unit 3, lights up sequentially in synchronization with the frame rate of the image sensors 51 and 52, so that the central wavelength of the object S being transported at a constant speed by the transport unit 2 is λ １ ~λ ５Inspection light L1 in each wavelength range is repeatedly irradiated in sequence. Furthermore, the light source unit 3A alternately switches between the output from light source devices 31A and 32A and the output from light source devices 31B and 32B. This alternately switches between a state where polarizers 93A and 97 function as orthogonal nicols with respect to the incident light L2, and a state where polarizers 93B and 97 function as parallel nicols with respect to the incident light L2. [Sixth Modification]

[0126] As shown in Figure 34, in the sixth modified example of the filter section 4C, a half-mirror 40H may be formed on the separation surface 41b. Also, in the filter section 4C, a band-pass filter or a step filter may be formed on the output surface 42c of the prism 42. In this case, in the filter section 4C, after the incident light L2 is separated by the half-mirror 40H, the image sensor 52 captures the first light that has passed through the band-pass filter or step filter on the output surface 42c. Also, after the incident light L2 is separated by the half-mirror 40H, the image sensor 51 captures the second light that is incident without passing through the band-pass filter or step filter. The first and second lights may be captured simultaneously. In this configuration, the image sensor 52 detects the first light, thereby determining the amount of transmitted light I Ｔ The light intensity I of the transmitted light is acquired, and the image sensor 51 detects the second light. Ｔ and the amount of reflected light I Ｒ The sum of these two amounts of light is effectively obtained. [Example of foreign object identification]

[0127] Figures 35 to 37 illustrate an example of the identification result of foreign matter W in the target object S by the identification unit 7. The target object S is PP (polypropylene), PE (polyethylene), and PS (polystyrene). In the example in Figures 35 to 37, the data generation unit 6 uses a central wavelength λ １ Let λ be 1200 nm, and the central wavelength λ ２ The wavelength was set to 1650 nm, and GC image data of the difference in light intensity was generated for each. In addition, in the examples of Figures 35 to 37, the data generation unit 6 has a center wavelength λ １ Light intensity and central wavelength λ ２ The ratio of wavelength to light intensity (I λ１ / I λ２The foreign substance W shall be identified using the following method. The central wavelength λ １ Light intensity I λ１ The central wavelength is λ １ Light intensity I Ｒ and light intensity I Ｔ It is the sum of and the central wavelength λ ２ Light intensity I λ２ The central wavelength is λ ２ Light intensity I Ｒ and light intensity I Ｔ It is the sum of and . Center wavelength λ １ Light intensity I λ１ and central wavelength λ ２ Light intensity I λ２ It is calculated by the following formula.

[0128] Figure 35(a) shows the identification results based solely on the wavelength intensity ratio. While PP, PE, and PS can be separated, the values ​​for the PE group and the PS group are close, so there is a risk of mixing depending on the error. Figure 35(b) shows the central wavelength λ １ This is an identification result based solely on the relative difference in light intensity. A mixture of PE and PS groups is present. Figure 35(c) shows the central wavelength λ ２ This identification result is based solely on the relative difference in light intensity. While the PP group is clearly separated from the other groups, the values ​​for the PE group and the PS group are close, so there is a risk of them being mixed together depending on the error.

[0129] Figure 36(a) shows the center wavelength λ １ The identification result is based on both the relative value of the difference in light intensity and the wavelength-light intensity ratio. Figure 36(b) shows the central wavelength λ ２ This is the identification result based on both the relative difference in light intensity and the wavelength-light intensity ratio. It can be seen that PP, PE, and PS are sufficiently separated. Figure 37 shows the central wavelength λ １ The relative value of the difference in light intensity and the central wavelength λ ２ This is the identification result based on both the relative value of the difference in light intensity and the wavelength ratio. It can be seen that PP, PE, and PS are all sufficiently separated. From the above results, the wavelength light intensity ratio and the central wavelength λ are used. １ The relative value of the difference in light intensity, and the central wavelength λ ２It can be seen that the identification accuracy of the identification unit 7 is improved by identifying the relative values ​​of the light intensity difference based on two or more parameters.

[0130] Examples of foreign object identification are not limited to those described above. The data generation unit 6 may obtain an approximate formula for the spectral reflectance spectrum of the object S based on the luminance data of transmitted light Lt and the luminance data of reflected light Lr. The identification unit 7 may then identify the material of the foreign object W or analyze its material composition based on the parameters of the approximate formula. Analysis of material composition may involve quantitative analysis of the material, such as analyzing properties of sugar, fat, water, protein, pH, etc. The photodetection unit 5B, configured as a multi-line sensor as shown in Figure 16, will be explained as an example.

[0131] The object S is irradiated with inspection light L1 having the reflection spectral characteristics shown in Figure 38. The central wavelength of this inspection light L1 is λ Ｍ = 1200 [nm]. The vertical axis of the graph in Figure 38 shows the relative radiant intensity, which is, for example, the value obtained by comparing the intensity of the inspection light L1 with the intensity of the reference light. When the image sensors 54 and 55 detect the incident light L2 reflected from the object S, an optical element 40C having the step-shaped filter characteristics shown in Figure 39 is placed in front of the image sensor 55. At this time, the values ​​of each parameter are calculated as follows from equations (9) to (12). J Ｔ０ =0.493436 J Ｒ０ =0.505745 J Ｔ１ =606.057 [nm] J Ｒ１ = 592.772 [nm] Also, the detected light quantities of image sensors 54 and 55 are I Ｔ +I Ｒ , I Ｔ It is represented by I Ｔ = 0.1106 I Ｒ = 0.1280 Furthermore, from equations (19) and (20), the slope k of the linear function approximation and the center wavelength λ Ｍ reflectance r Ｍ It is calculated as follows: k = -0.0005183 [nm] －１ ] r Ｍ = 0.23871 Figure 41 shows the derived linear function approximation r ０１(λ) is shown. For example, approximate r ０１ The parameters of (λ) are the slope k and the center wavelength λ. Ｍ reflectance r Ｍ Based on this, the material of the foreign object W may be identified or its material composition may be analyzed.

[0132] When the image sensors 54 and 56 detect the incident light L2 reflected from the object S, an optical element 40D having the bandpass filter characteristics shown in Figure 40 is placed in front of the image sensor 56. At this time, the values ​​of each parameter are calculated as follows from equations (27) to (32). Ｔ０ = 0.69244 J Ｒ０ = 0.30674 J Ｔ１ =831.081 [nm] J Ｒ１ =367.748 [nm] J Ｔ２ =997738 [nm ２ ] J Ｒ２ =441880[nm ２ Furthermore, the detected light quantities of the image sensors 54 and 56 are each I Ｔ +I Ｒ , I Ｔ It is represented by I Ｔ =0.15323, I Ｒ = 0.08538. Furthermore, from equations (40) and (41), the degree of curvature a and the center wavelength λ of the quadratic function approximation formula can be obtained. Ｍ reflectance r Ｍ It is calculated as follows: a = 0.0000198959 [nm] －２ ] r Ｍ = 0.045436 Figure 41 shows the derived quadratic function approximation r ０２ (λ) is shown. For example, approximate r ０２ The parameters of (λ) are the degree of curvature a and the center wavelength λ Ｍ reflectance r Ｍ Based on this, the material of the foreign object W may be identified or its material composition may be analyzed.

[0133] Furthermore, the linear function approximation r ０１ (λ) and the quadratic function approximation r ０２ When (λ) is combined, the approximate r shown in Figure 41 is obtained. ０１２(λ) Here, the approximation r is a combination of a linear function and a quadratic function. ０１２ (λ) is expressed as follows: Here, r Ｍ12 r was obtained using a linear function approximation. Ｍ The value of r obtained by quadratic function approximation Ｍ This is the average value of the values.

[0134] The spectrum S2 shown in Figure 41 is an example of the spectral reflectance spectrum of an actual object S. Approximate r ０１２ (λ) confirms that the approximation of spectrum S2 holds true. This approximation may be used to identify the material of foreign matter W or to analyze its material composition.

[0135] 1, 1A, 1B, 1C, 1D, 1E, 1F... Inspection device, 3, 3A... Light source unit, 4, 4A, 4B, 4C... Filter unit, 5, 5A, 5B, 5C, 5D... Light detection unit, 6... Data generation unit, 7... Identification unit, 40, 40A... Optical element (first optical element), 40B... Optical element (second optical element), 51... Image sensor (first image sensor), 52... Image sensor (second image sensor), 54... Image sensor (first image sensor), 55... Image sensor (second image sensor), 93, 93A, 93B... Polarizing plate, 92... Polarization rotation unit, BS... Polarizing beam splitter, D1, D2... Polarization direction, GC... Light intensity difference image data, GL... Light intensity ratio image data, GR... Specular reflection image data, I Ｒ , I Ｔ ...light intensity, L1...inspection light, L2...incident light, Lr...reflected light, Lt, Lt1, Lt2, Ltp, Lts...transmitted light, Pt...transparent body, S...object, S01...output step, S02...light guiding step, S05...identification step, W...foreign object, λ Ｍ , λ １ ~λ ５ ,...center wavelength.

Claims

1. An inspection apparatus comprising: a light source unit that outputs inspection light including a wavelength distribution in a predetermined wavelength range toward an object; a filter unit including a first optical element having a step-like filter characteristic in which the transmittance of light changes in a step-like manner at a first edge wavelength within the predetermined wavelength range; a light detection unit that receives a first light, which is either transmitted light that has passed through the first optical element or reflected light that has been reflected by the first optical element, and a second light, which is at least the other of the transmitted light and the reflected light, from the incident light that the inspection light has reflected or passed through the object, and outputs luminance data of the first light and luminance data of the second light; a data generation unit that generates characteristic data for the incident light from the object using the luminance data of the transmitted light and the luminance data of the reflected light based on the luminance data of the first light and the luminance data of the second light; and an identification unit that identifies the material of a foreign object in the object based on the characteristic data.

2. The inspection apparatus according to claim 1, wherein the first edge wavelength is located within the full width at half maximum of the wavelength distribution of the inspection light.

3. The inspection apparatus according to claim 1 or 2, wherein the light detection unit comprises a first image sensor for receiving reflected light and a second image sensor for receiving transmitted light.

4. The inspection apparatus according to any one of claims 1 to 3, wherein the first edge wavelength is the central wavelength of the inspection light.

5. The inspection apparatus according to any one of claims 1 to 3, wherein the first edge wavelength includes a second edge wavelength at which the transmittance of the light changes in a step-like manner between a minimum value and a maximum value, and a third edge wavelength that is larger than the second edge wavelength at which the transmittance of the light changes in a step-like manner between a maximum value and a minimum value in the opposite direction to the direction of change in the transmittance at the second edge wavelength.

6. The inspection apparatus according to any one of claims 1 to 5, wherein the first optical element has the step-shaped filter characteristics in each of a plurality of wavelength regions that are different from each other.

7. The inspection apparatus according to any one of claims 1 to 6, wherein the filter section further comprises a second optical element having the step-shaped filter characteristics at a fourth edge wavelength that is different from the first edge wavelength.

8. The inspection apparatus according to claim 1 or 2, wherein the light detection unit is configured such that a plurality of image sensors, each including an image sensor that receives transmitted light transmitted through the first optical element by a plurality of pixels arranged in one direction, are arranged along a direction intersecting the one direction.

9. The inspection apparatus according to any one of claims 1 to 8, further comprising a polarization rotation unit disposed before the filter unit in the optical path of the incident light, which rotates the polarization direction of the incident light.

10. An inspection apparatus according to any one of claims 1 to 9, further comprising a polarizing plate positioned downstream of the light source in the optical path of the inspection light, wherein the filter unit further includes a polarizing beam splitter that reflects one of the incident light's polarization component parallel to the transmission axis of the polarizing plate and the polarization component perpendicular to the transmission axis of the polarizing plate, and transmits the other of the incident light's polarization component parallel to the transmission axis of the polarizing plate and the polarization component perpendicular to the transmission axis of the polarizing plate, or a polarizing plate that blocks one of the polarization components and transmits the other of the polarization components.

11. The inspection apparatus according to any one of claims 1 to 10, wherein the data generation unit generates specular reflection image data based on the characteristic data, and the identification unit determines whether or not the object contains a transparent object based on a comparison between the brightness of the specular reflection image data and a threshold.

12. The inspection apparatus according to any one of claims 1 to 11, wherein the data generation unit generates, as characteristic data, light intensity ratio image data showing the ratio of the light intensity of transmitted light to the light intensity of reflected light, or light intensity difference image data based on the difference between the light intensity of transmitted light and the light intensity of reflected light, and the identification unit identifies the material of the foreign substance in the object based on either the light intensity ratio image data or the light intensity difference image data.

13. The inspection apparatus according to any one of claims 1 to 12, wherein the data generation unit obtains an approximate formula for the spectral reflectance spectrum of the object based on the luminance data of the transmitted light and the luminance data of the reflected light, and the identification unit performs identification of the material of the foreign object or analysis of the material composition based on the parameters of the approximate formula.

14. The inspection apparatus according to claim 13, wherein the parameters are parameters based on the slope of the linear function approximation formula and the reflectance at the central wavelength.

15. The inspection apparatus according to claim 13, wherein the parameters are parameters based on the degree of curvature of the quadratic function approximation formula and the reflectance at the central wavelength.

16. An inspection method comprising: an output step of outputting inspection light including a wavelength distribution in a predetermined wavelength range toward an object; a light guiding step of guiding incident light generated by the reflection or transmission of the inspection light at the object to a first optical element having step-like filter characteristics in which the transmittance of light changes in a step-like manner at a first edge wavelength within the predetermined wavelength range; a light detection step of receiving a first light from the incident light, which is either transmitted light that has passed through the first optical element or reflected light that has been reflected by the first optical element, and a second light that includes at least the other of the transmitted light and the reflected light, and outputting luminance data of the first light and luminance data of the second light; a generation step of generating characteristic data for the incident light from the object using the luminance data of the transmitted light and the luminance data of the reflected light based on the luminance data of the first light and the luminance data of the second light; and an identification step of identifying the material of a foreign substance in the object based on the characteristic data.

Images