Optical Inspection System for Detecting Light Path Deviation and its Inspection Method

Abstract

An optical inspection system for inspecting surface defects of a non-flat specimen according to the invention comprises: an illumination unit that irradiates the specimen with a single light source; an object lens that receives light reflected or refracted from the specimen to form a primary image; a relay lens that adjusts the optical path as needed; a tube lens that forms a final image; a pinhole that passes only light within an allowable angle (μ) range; and a sensor that receives light passing through the pinhole to form an image; and is configured to position the entire specimen within the focal range by increasing the F / # value of the optical inspection system to expand the depth of field.

Description

Technology Field

[0001] The present invention relates to an optical inspection system for detecting path deviation of light and an inspection method thereof. Furthermore, the present invention also relates to an aperture-based imaging system for detecting path deviation of illumination light based on angle and a control method thereof. Background Technology

[0002] An optical inspection system is composed of a light source (illumination), an imaging optical system, a camera, etc., and such a general imaging optical system is configured to form an image of an object by controlling light reflected or emitted from the object through optical elements such as lenses or mirrors.

[0003] Conventional two-dimensional optical inspection devices are primarily designed and operated for flat specimens such as thin films, glass, flexible printed circuit boards (FPCB), display substrates, and semiconductor wafers. However, there is a growing demand for optical inspection of non-planar specimens in industrial settings.

[0004] The optical and illumination systems include various forms of illumination depending on the object being inspected, and consist of lenses and sensors for forming an image after the irradiated light is reflected from the specimen.

[0005] In the case of non-flat specimens, the focal plane of the imaging system is flat and may not coincide with the object being inspected; accordingly, a device to shift the focal plane is sometimes added. To mitigate this problem, some conventional devices introduce multi-angle lighting, lighting split control, and designs to increase depth of field (increasing F / #, i.e., focal length / aperture diameter, or applying apochromatic lenses).

[0006] In terms of sensors and scanning methods, conventional line scanning methods achieve optimal performance in structures where the specimen is transported in a straight line; however, on curved or irregular surfaces, it is difficult to maintain optical focus, which can lead to unstable defect detection.

[0007] For curved specimens, area sensors are used instead of line scanning, or inspection systems with autofocus tracking are utilized after acquiring shape information of the specimen in advance. However, many 2D inspection systems are still optimized for flat surfaces, so problems such as luminance distortion, distortion, and reduced defect detection rates due to curvature and surface inclination exist.

[0008] In particular, conventional technology, in which most specimens are flat, comprises an optical system, an illumination system, a sensor, and image processing, and has the following limitations regarding curved specimens.

[0009] First, in specimens with irregular curvature, the angle and intensity of reflected light are inconsistent, which reduces inspection accuracy, and

[0010] Second, due to differences in focal length caused by surface geometry, it is difficult to maintain uniform focus across the entire line in the line scanning method, resulting in blurring.

[0011] Third, also, since the lighting area and sensor field of view are fixed, in the tilted area, saturation occurs due to insufficient light irradiation or strong reflection, and

[0012] Fourth, the lack of inspection algorithms that reflect curved shapes increases the false positive rate in image matching or pattern-based defect detection.

[0013] Fifth, furthermore, since non-flat specimens are sensitive to rotation, vibration, and minute positional errors during the inspection process, these factors act as serious error causes in high-precision roll-to-roll inspection systems.

[0014] General two-dimensional optical inspection systems have been optimized for flat specimens such as thin films, glass, FPCBs, display substrates, and semiconductor wafers.

[0015] However, in industrial settings, the demand for inspecting curved or irregular specimens is increasing; in such cases, maintaining focus is difficult due to local variations in lighting angles and reflection conditions, leading to an increase in false positives or missed detections.

[0016] Increasing the F-number to enhance depth of field, multi-angle lighting, and autofocus tracking increase system complexity and costs while slowing down inspection speed.

[0017] While the confocal method suppresses signals outside the depth of focus through a pinhole, its primary objective is to improve axial resolution; therefore, there are limitations in quantitatively controlling sensitivity to path angle variations using specific angle parameters. Consequently, a simple optical structure is required that selectively passes only the normal path based on the angle variation of the optical path.

[0018] Therefore, an optical system is required that can rapidly and reliably detect cases where the path of illumination light is deformed by external factors such as non-flat specimens, surface inclination, refractive index discontinuities, foreign matter, or scratches. Furthermore, it is necessary to parameterize sensitivity so that it can be easily set to optimal values ​​for each product process. Prior art literature

[0019] Japanese Registered Patent No. 5087186 The problem to be solved

[0020] Therefore, there is a need for an optical inspection system that detects light path deviations and an inspection method thereof to solve these problems. means of solving the problem

[0021] The optical inspection system for inspecting surface defects of a non-flat specimen according to the present invention, which solves the above-mentioned problem, comprises: an illumination unit that irradiates the specimen with a single light source; an object lens that receives light reflected or refracted from the specimen to form a primary image; a relay lens that adjusts the optical path when necessary; a tube lens that forms a final image; a pinhole that passes only light within an allowable angle (μ) range; and a sensor that receives light passing through the pinhole to form an image; and is configured to position the entire specimen within the focal range by increasing the F / # value of the optical inspection system to expand the depth of field.

[0022] In the optical inspection system of the present invention, the single light source provides a diffused beam, and the pinhole may be configured to allow only normal light rays to pass through and to block light rays whose paths have been displaced by reflection, refraction, shielding, etc.

[0023] The optical inspection system of the present invention may be configured to distinguish between normal and abnormal rays based on the allowable angle μ at the pinhole position, and to block rays exceeding μ.

[0024] In the optical inspection system of the present invention, an aperture stop may be additionally placed on the optical path to limit the maximum aperture and angle of incidence of the incident light.

[0025] The optical inspection system of the present invention can be configured to selectively pass only light rays within a μ range based on the local height difference (h) and local inclination (θ) of the specimen relative to the reference plane.

[0026] In the optical inspection system of the present invention, the size of the pinhole can be determined by setting the boundary between the normal light ray and the abnormal light ray to an angle μ.

[0027] The optical inspection method for inspecting surface defects of a non-flat specimen according to the present invention comprises the steps of: irradiating a specimen with a single light source; receiving light reflected or refracted from the specimen with an object lens to form a primary image; adjusting the optical path with a relay lens if necessary; performing a final image with a tube lens; passing light through a pinhole to allow only light within an allowable angle (μ) range to pass through; and receiving the light that has passed through the pinhole with a sensor to form an image; and is configured to increase the F / # value of the optical inspection system to expand the depth of field.

[0028] In the optical inspection method of the present invention, the single light source provides a diffused beam, and the pinhole may be configured to allow only normal light rays to pass through and to block light rays whose paths have been displaced by reflection, refraction, shielding, etc.

[0029] The optical inspection method of the present invention may be configured to include the step of distinguishing between normal rays and abnormal rays based on an allowable angle μ at the pinhole location, and blocking rays exceeding μ.

[0030] The optical inspection method of the present invention may be configured to further include the step of limiting the maximum aperture and angle of incidence of incident light by additionally placing an aperture on the optical path.

[0031] The optical inspection method of the present invention may be configured to include a step of selectively passing only light rays within a μ range based on the local height difference and local inclination of the specimen relative to a reference plane.

[0032] In the optical inspection method of the present invention, the size of the pinhole may be configured to be determined by setting the boundary between the normal light ray and the abnormal light ray to an angle μ. Effects of the invention

[0033] By employing the optical system for detecting light path deviation and the control method provided according to the present invention, the constraints on the position, size, and shape of inspection targets that were limited due to the depth of focus limitations of conventional optical systems can be improved, which is advantageous in terms of expanding inspection targets.

[0034] In addition, conventionally, a scanning method including distance variation was mainly used to overcome these limitations, but this resulted in an increase in inspection time; whereas, in the present invention, the inspection speed of the entire system can be improved by omitting scanning in the distance direction.

[0035] In addition, while existing 2D optical inspection devices are optimized for flat objects such as line scans and area image sensors, distortion and defect detection errors frequently occur in non-flat (curved) specimens due to changes in the angle and intensity of reflected light and differences in focal length. In contrast, the present invention applies the optimization of the F / # design of the imaging optical system and the selective optical path blocking technology for pinholes together, thereby enabling stable and precise defect detection even for curved and irregular specimens, and significantly reducing image brightness distortion, focus blur, and false detection rates.

[0036] As a result, it is possible to achieve a significant reduction in the inspection error rate for various components having fine curvature and inclination, such as industrial films, semiconductors, and display substrates.

[0037] In addition, there may be other advantages and effects of the present invention that are not mentioned above but can be realized by the means and combinations thereof set forth in the patent claims. Brief explanation of the drawing

[0038] Figure 1 is a configuration diagram showing the system of the present invention based on an infinite light source. FIG. 2 is a drawing showing a first embodiment of an example of optical path variation of the present invention. FIG. 3 is a drawing showing a second embodiment of an example of optical path variation of the present invention. FIG. 4 is a drawing showing a third embodiment of an example of optical path variation of the present invention. FIG. 5 is a drawing showing a fourth embodiment of an example of optical path variation of the present invention. FIG. 6 is a diagram illustrating a structure in which only light rays traveling at an angle within μ are allowed to pass through in the optical system of the present invention, and light rays exceeding μ are blocked. Figure 7 is a diagram showing a comparison between the case where light irradiated from a light source in the present invention forms an image on a sensor along a normal path and the case where it deviates from the path and is blocked. Figure 8 is a diagram showing the path of light passing through the pinhole of the present invention. Figure 9 is a diagram showing the entire optical path through which light from an object passes through several lenses and apertures to reach the image plane in an optical system. Specific details for implementing the invention

[0039] The present invention will be described in more detail below.

[0040] The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform a person skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims.

[0041] In this specification, the singular form includes the plural form unless specifically stated otherwise in the text. Unless otherwise defined, all terms used in this specification (including technical and scientific terms) may be used in a meaning commonly understood by a person skilled in the art. Furthermore, terms defined in advance are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise.

[0042] The following embodiments are examples for explaining the present invention and are not limited thereto; the present invention can be modified in any way within the scope described in the claims.

[0043] The present invention relates to an optical inspection system and a method capable of inspecting surface defects of a non-flat specimen with high resolution and high contrast. More specifically, the invention relates to a technology that suppresses focus blur and contrast degradation caused by the curvature, tilt, height difference, etc. of the specimen, and expands the depth of field to provide homogeneous image quality across the entire specimen.

[0044] To describe the configuration of the system, an optical inspection system according to one embodiment of the present invention includes an illumination unit that irradiates a specimen with a single light source, an object lens that receives light reflected or refracted from the specimen to form a primary image, a relay lens that adjusts the optical path when necessary, a tube lens that forms a final image, a pinhole that allows only light within an allowable angle (μ) range to pass through, and a sensor that receives light passing through the pinhole to form an image.

[0045] The above optical inspection system is designed to increase the F / # value to expand the depth of field, thereby enabling imaging of areas of the specimen with various heights and inclinations within the same focal range.

[0046] Regarding the light source conditions, the present invention uses a diffused beam as the light source and does not use a collimated beam, a single LED, or a non-diffusing beam. This is intended to ensure uniform illumination across the entire specimen area and to maximize contrast enhancement in combination with a light blocking function based on angle selectivity (μ).

[0047] Regarding the pinhole and angle selection, the pinhole is placed at a specific location on the optical system and distinguishes between normal path and abnormal path rays based on the allowable angle μ at that location. Light incident at an angle exceeding μ is blocked by the pinhole, and only rays within the μ range reach the sensor. This prevents focus blur and contrast degradation caused by scattered light that occur in areas with severe curvature or large slopes.

[0048] To explain aperture and angle of incidence control, an aperture stop positioned in the optical path limits the maximum aperture of the incident light and blocks the influx of unwanted ambient light. The aperture and pinhole operate complementarily to selectively allow only light rays within a specific angle range to pass through.

[0049] Regarding the selection based on height difference and inclination, the present invention measures or estimates the local height difference (h) and local inclination (θ) of a specimen relative to a reference plane, and is designed to selectively allow only light rays falling within the μ range to pass through. For example, light rays exceeding μ are blocked, thereby eliminating unnecessary scattered light or background light.

[0050] In addition, the optical inspection method of the present invention comprises: (a) irradiating a specimen with a single light source; (b) receiving light reflected or refracted from the specimen with an object lens to form a primary image; (c) adjusting the optical path with a relay lens if necessary; (d) performing a final image with a tube lens; (e) allowing only light within the μ range to pass through a pinhole; and (f) forming an image with a sensor.

[0051] The present invention is a system that images only light whose path has not changed, and if light is refracted due to any cause, that light is blocked by a pinhole aperture.

[0052] The lighting unit of the present invention is composed of a single light source regardless of position and focal length, and lighting in the form of a collimated beam, a single LED, or a not-diffusing beam is not used. In the following description, the collimated beam is used as the reference.

[0053] The expression "based on an infinite light source" as used in this specification should be understood not as meaning that the infinite light source itself is used in an actual device, but rather that the optical design, optical path description, and imaging principle description are simplified and described under "infinite light source conditions." In other words, the infinite light source (collimated beam) is not used as an actual lighting unit; rather, it assumes the ideal condition that light enters as "parallel light" when explaining the operating principles of an optical system or showing how light rays propagate in drawings. This is intended to exclude changes in the optical path due to the type of lighting, diffusion, LED characteristics, etc., and to clearly explain lens placement, focus relationships, pinhole effects, etc.

[0054] Therefore, it should be understood that other forms (e.g., diffuse lighting, multiple LEDs, lasers, etc.) can be used in the actual lighting section.

[0055] The imaging optical system may include an objective lens, which is the first lens group that receives light emitted from an object of observation to form an image, a relay lens that can be optionally adopted to avoid spatial constraints in system configuration, and a tube lens for final imaging.

[0056] A pinhole is an aperture that allows only light that has been altered within an acceptable range—that is, light that has not been altered by the imaging target—to pass through. Pure illumination light is focused and formed at the position of a dot aperture, passing through the pinhole placed at that location.

[0057] The sensor unit receives light passing through the pinhole and converts it into an electrical signal.

[0058] In the present invention, the term "normal path" refers to light whose direction of light from the lighting unit is not changed by external interference, and external interference refers to a change in the path of light caused by refraction and reflection of light due to foreign substances, obstacles, differences in medium, etc.

[0059] Figure 1 is a configuration diagram showing the system of the present invention based on an infinite light source. Referring to this figure, the optical system according to the present invention is divided into an illumination unit and an imaging unit, and each unit is composed of a plurality of optical elements that perform specific functions and roles.

[0060] In one embodiment of the present invention, an illumination system composed of a single light source is used.

[0061] The illumination unit consists of a light source and a collimator (EFL-1), which converts light emitted from the light source into parallel light and illuminates the object of observation uniformly. Light passing through the collimator enters the objective lens (EFL-2), which belongs to the imaging unit; this objective lens collects light reflected or transmitted from the object and transmits it to the imaging unit. In other words, the objective lens receives light rays reflected or refracted from the object to form the first image of the optical system.

[0062] The light path formed in the illumination unit includes both signal light, including whether it is modulated by the target object, and unmodulated background light.

[0063] The imaging unit consists of the above-mentioned objective lens (EFL-2), relay lens-1 (EFL-3), relay lens-2 (EFL-4), pinhole, and tube lens (EFL-5). Relay lens-1 receives light passing through the objective lens and forms an intermediate image, and then relay lens-2 repositions it to focus the light rays at the pinhole position.

[0064] A pinhole allows light that is not modulated by the imaging target (within the allowable range) to pass through, while effectively blocking modulated, unnecessary scattered light or noise light. This process contributes to the removal of background light and the improvement of the signal-to-noise ratio (SNR). In other words, among the light emitted from the imaging target (object), the pinhole selectively allows only light whose optical path has not changed (within the allowable displacement) to pass through. In this invention, only normal light rays pass through the pinhole on the optical axis, which serves as an exemplary reference axis, while light whose path has been altered due to refraction, reflection, etc., is blocked at the pinhole.

[0065] Light passing through the pinhole is finally formed by the tube lens (EFL-5) and reaches the camera sensor. The tube lens plays a crucial role in determining the image magnification ratio and resolution, and is a key element in adjusting the focal length and magnification of the entire optical system. The tube lens is responsible for final image formation and transmits the image to the sensor.

[0066] The sensor detects only the light that has passed through the pinhole, converts it into image information, and finally, the sensor obtains a high-resolution, high-precision image in which light rays from abnormal paths are excluded.

[0067] The distance between each optical element is determined by the Effective Focal Length (EFL) value, which directly affects the total magnification, aberration correction, depth of focus, and field of view of the optical system. The optical design of the present invention is optimized to maintain focus even on non-planar specimens and enables high-precision defect detection through background light suppression using pinholes.

[0068] In other words, this system has a structure in which light emitted from a light source is focused by an objective lens after passing through a collimator, and then forms a high-resolution image by a tube lens after passing through a relay lens group and a pinhole; through this, stable and precise optical inspection is possible even for specimens with non-flat surfaces, curved surfaces, or irregular shapes.

[0069] Meanwhile, the present invention applies a design principle of adjusting the F / # value of the optical system to maximize inspection efficiency and accuracy for non-flat specimens (curved or irregular specimens). The F / # (F-number) can be defined as the value obtained by dividing the focal length (f) of the lens by the aperture diameter (D) (F / # = f / D). Increasing the F / # (reducing the aperture) reduces the amount of light incident on the lens while simultaneously widening the depth of field (DOF). In particular, for non-flat specimens with varying curvatures and inclinations, it is difficult to keep the entire inspection surface within focus; however, increasing the F / # value reduces the degree to which the optical system opens, allowing a wider area to be clearly imaged simultaneously. Consequently, problems with maintaining focus (blurring) and inspection errors caused by variations in the specimen's surface shape, inclination, or rotation are significantly reduced.

[0070] However, since the amount of light is reduced, there are limitations such as increased noise, decreased sensitivity, and longer exposure times; nevertheless, increasing the F / # can be advantageous when high-resolution inspection is required.

[0071] In addition, even if the surface curvature, microstructure, and inclination of the non-flat specimens fluctuate during inspection, the in-focus area within the same optical system can be maximized by utilizing the magnified depth of field.

[0072] Therefore, by adopting the present invention, it becomes suitable for inspecting various specimens such as non-flat films, curved semiconductors, and display modules.

[0073] The following description refers to FIGS. 2 to 5, which show exemplary embodiments of the optical path variation of the present invention.

[0074] FIG. 2 illustrates an example where illumination light passing through a flat objective lens is locally scattered by fine defects (particles, scratches, voids, etc.) on the imaging target, causing the optical path to shift at multiple angles. The parallel beams on the left proceed in different directions (red arrows) on the right after passing through the defect, which is observed as an effect where the position spreads out radially from the center point in the Fourier or conjugate plane of the imaging system. Since the pinhole of the present invention allows only normal light rays within an allowable half-angle μ to pass through, light rays that have undergone angular shifting due to the defect as shown in FIG. 2 are mostly blocked at the pinhole, and only the normal path component reaches the sensor. This significantly improves contrast around the defect.

[0075] FIG. 3 illustrates optical path variation caused by a rough or non-convex surface. When a relative height difference h with respect to a reference plane and a local slope θ exist, light rays emitted from various points within or outside Abbe's compensation error limit are deflected at different angles, and as a result, scattered light rays of various angles are formed in the right-hand region. In this state, even at the same location, the emission angle distribution varies depending on the combination of surface height and slope, so the energy distribution in the Fourier or conjugate plane spreads widely. The pinhole of the present invention selectively removes components outside the angle acceptance range defined by μ, thereby effectively suppressing AC error scattering components originating from non-convex surfaces.

[0076] FIG. 4 illustrates a case where the entire beam is collectively tilted by a certain angle θ due to a parallelism error of the test specimen or optical element relative to a reference plane. The parallel beam bundle on the left is deflected by a certain angle within the parallelism error limit and propagates to the right, and the acceptable half-angle μ is indicated in the figure. In this system, since the center of the pinhole on the Fourier or conjugate plane is aligned with the normal path, if the collective tilt of the beam exceeds μ, the center energy moves out of the pinhole opening, causing a rapid reduction in the signal. Therefore, FIG. 4 is an example showing that the system responds sensitively to parallelism error in an angle-selective manner, suggesting that an acceptable tilt threshold angle can be process-defined through the set value of μ.

[0077] Figure 5 illustrates a case where a defect directly occludes light, resulting in the loss of specific light components. After passing through the objective lens, the parallel beam of light on the left has a portion of its optical path physically blocked by an opaque defect inside or on the surface of the test specimen; consequently, a specific strip of the beam of light on the right is lost. In this case, the main change is expressed as a localized decrease in intensity rather than an angle deviation, and while the normal path component passing through the pinhole passes through, the total amount of light is partially reduced due to the occlusion. In this system, such occluded defects are also clearly expressed as intensity non-uniformity on the sensor surface, which is advantageous for defect detection.

[0078] Figures 2 and 3 are similar in that they both present "angular displacement type" defect cases, but while Figure 2 emphasizes a random scattering type angular distribution caused by local discontinuities (particles, scratches, etc.), Figure 3 shows a situation where a position-dependent and shape-induced angular distribution spreads widely due to a combination of surface roughness, height difference, and inclination. In both cases, a significant portion of the scattering component is excluded through the half-angle μ of the pinhole, thereby improving contrast, and there is a design trade-off in that the scattering suppression effect increases as μ becomes smaller (the sensitivity increases), but the light loss increases.

[0079] Figure 4 shows a parallelism error case where the entire beam is tilted at a constant angle, and unlike the "dispersive scattering" of Figures 2 and 3, the center of energy shifts uniformly. In this case, if the center point in the Fourier plane or conjugate plane deviates to the outside of the pinhole opening, the signal drops sharply; therefore, setting μ is equivalent to determining the allowable tilt value (θ threshold) permitted by the system. In applications where parallelism management is important during the process, the mechanism of Figure 4 can be used to quantify the allowable angle and determine equipment alignment.

[0080] Figure 5 is a typical example of a shielded defect, in which intensity loss is dominant rather than the broadening of the angular spectrum. Since the pinhole allows light to pass through the normal path, regions with small angular deviations pass through as is, but due to shielding, specific groups of light are lost, appearing as linear or localized photosensitive patterns on the sensor surface. Therefore, the key aspect of Figure 5 is the analysis of intensity non-uniformity on the imaging surface rather than the angular selectivity of the pinhole, and it is used to distinguish defect types in a complementary manner with Figures 2 to 4.

[0081] In summary, FIG. 2 illustrates local defect-induced scattering, FIG. 3 illustrates position-dependent scattering combined with surface shape and AC error, FIG. 4 illustrates collective angular displacement due to parallelism error, and FIG. 5 illustrates intensity loss due to shielding. The pinhole-based angle-selective imaging structure of the present invention provides high contrast for angular displacement type defects corresponding to FIG. 2 to 4, clearly reveals defects due to local non-uniformity of the intensity profile for shielding type defects corresponding to FIG. 5, and can stably detect four types of defects by mutually distinguishing them by combining the set value of μ with lighting and exposure conditions.

[0082] In this regard, the present invention performs angle-selective light reception using a pinhole placed on a Fourier or conjugate plane, thereby selectively suppressing "angle-displacement" components such as local defect-induced scattering as in FIG. 2, position-dependent scattering due to a combination of surface roughness, height difference, and tilt as in FIG. 3, and collective beam tilt due to parallelism error as in FIG. 4, and allowing only the normal path components to pass through. As a result, defect contrast is significantly increased and noise caused by background light and diffuse reflection is reduced.

[0083] In addition, sensitivity can be quantitatively designed and operated by introducing an allowable half-angle μ as a parameter. Since μ is defined as a function of the pinhole aperture and the effective focal length, the allowable critical angle can be numerically defined for the angular displacement distribution shown in FIGS. 2 to 4, and a reproducible calibration standard between devices is established. Reducing μ strengthens the suppression of scattering components and increases contrast against fine defects, while increasing μ is advantageous for securing light intensity and processing speed, allowing the selection of an optimal point tailored to the process conditions.

[0084] Angle-selective filtering enables stable imaging even on non-planar specimens. As shown in Fig. 3, even when the emission angle distribution widens due to the surface shape, only normal components satisfying the μ condition are acquired, so light intensity fluctuations and distortion caused by curvature or local tilt are mitigated, and the reliability of the defect signal is improved due to angle-based separation even if the focal position varies slightly.

[0085] Meanwhile, the inspection speed can be increased by reducing the dependency on scanning in the distance (z) direction. Conventional methods performed stepwise or continuous z-scans to compensate for the problem of maintaining focus on curved or irregular specimens, but the present invention can secure effective contrast by suppressing only the angular displacement shown in FIGS. 2 to 4, thereby enabling a reduction or omission of the number of z-scans and improving throughput.

[0086] In addition, the signal-to-noise ratio (SNR) and dynamic range are expanded. Since the pinhole blocks components of angles other than μ, unnecessary slope components are naturally suppressed in the situation of Fig. 4 where the center of energy moves away from the pinhole opening, and the diffuse reflection background of Figs. 2 and 3 is reduced. As a result, more of the sensor's effective bit width can be allocated to the defect signal, and the detection lower limit is lowered.

[0087] Furthermore, defect types are easily separated. Since shielding type (intensity defect) defects, such as those in FIG. 5, are represented as local defects in the intensity profile rather than the angle spectrum, their representation on the sensor surface differs from that of angle displacement type defects (Figs. 2 to 4). The present invention can simultaneously reduce false positives and false negatives by separating two series of signatures through a μ sweep (variable aperture) or multiple exposures.

[0088] By adopting the present invention, process alignment and device health checks are quantified. As shown in FIG. 4, if a parallelism error exists, the energy center of the polyhedron deviates from the pinhole axis; therefore, the equipment alignment status can be numerically determined by measuring the acceptance half-angle and the center displacement. This provides a repeatable reference indicator during the installation and maintenance process.

[0089] Furthermore, it exhibits high complementarity with depth-of-field expansion designs. By combining a design that expands depth of field through increasing F / # with μ parameter control, focus retention (depth of field) and angle selection (contrast) can be achieved simultaneously on non-planar specimens. This ensures consistent quality over a wider process area compared to a single approach and allows for the simplification of the optical configuration. Relay lenses are selectively adopted based on spatial constraints, and since the key element is the pinhole placement in the Fourier or conjugate planes, high-contrast inspection is possible without complex multi-angle lighting or expensive scanning modules. This increases reliability while lowering system costs and alignment difficulty.

[0090] The present invention can also be universally applied to various specimens, such as reflective and transmissive types, and flat and non-flat types. FIGS. 2 to 5 cover major optical path disturbance factors such as scattering, shape-induced displacement, tilt error, and shielding, and since the angle selection structure of the present invention operates in common with these factors, consistency in principle is maintained even when the target and lighting conditions change.

[0091] Furthermore, adopting the present invention provides robustness against movement and vibration in actual usage environments. Even if micro-vibrations or positional errors occur in the specimen, abnormal path components deviating from the μ condition are filtered out, thereby mitigating the blurring and alignment instability commonly observed in line scan-based systems.

[0092] In summary, the technical advantages of the present invention lie in structurally eliminating signal turbidity caused by diffuse reflection, parallelism errors, scattering, and shielding in non-flat and curved specimens through angle-selective light reception quantified in μ, thereby simultaneously ensuring speed and reliability by reducing scan dependency, and providing a calibrated and highly process-portable inspection platform. These characteristics can be extended not only to industrial films, semiconductors, and displays but also to specimens for medical devices with significant surface shape variations.

[0093] FIG. 6 illustrates the principle in which a pinhole positioned on a Fourier (or its conjugate) plane in the imaging optical system of the present invention selectively passes or blocks light rays according to an angle acceptance half-angle (μ), and the process in which this principle is imaged to a final camera sensor through relay lens groups (EFL-3, EFL-4) and a tube lens (EFL-5). The objective lens (Objective, EFL-2) on the left collects light emitted from a specimen and transmits it to the imaging system, and relay lens-1 and relay lens-2 are positioned to precisely transmit the pommel (or space containing angle information) of the objective lens to the Fourier / conjugate plane where the pinhole is located. The pinhole is denoted by an aperture radius h, and this aperture allows only regular path light rays within an allowable half-angle μ with respect to the optical axis, which is an exemplary reference axis, to pass through, and blocks path deviation light rays exceeding μ. The light passing through the pinhole is formed into a final image by the tube lens (EFL-5) and reaches the camera sensor.

[0094] For reference, the reference axis on which apertures such as point apertures or pinholes are located has been described below as an "optical axis" or "optical axis" as an example. However, this is merely an example, and the reference axis defining the placement position of the aperture may be arbitrarily set according to the system's requirements, in addition to the optical axis. Furthermore, the allowable angle (e.g., μ) and relative height (e.g., h) determined based on the reference axis have also been described based on the optical axis for the convenience of explanation; however, this is merely intended to aid in the intuitive understanding of the invention through the drawings and should not be interpreted as limiting the scope of the present invention to a specific reference axis (e.g., optical axis).

[0095] In addition, in Fig. 6, among the beam of rays passing through the objective lens, the component traveling at a small angle around the optical axis is the regular path within μ, and only this component passes through the pinhole. Around the pinhole in the upper right of Fig. 6, two regions are distinguished by color. For convenience of explanation, the left wedge-shaped region with a green border conceptually represents the path within μ (passage target), and the red wedge-shaped region represents the path exceeding μ (blockage target). In an actual device, μ is determined by the pinhole aperture h and the effective focal length feq of the imaging system, and is based on the following formula:

[0096]

[0097] Here, h is the pinhole aperture radius, feq is the effective focal length, and μ is the reception half-angle.

[0098] feq is an effective value determined by the arrangement of EFL-2, EFL-3, EFL-4, and EFL-5, and the position and scale of the poly / conjugate plane are corrected according to the relay configuration (e.g., 4f or variant 4f).

[0099] The system flow at the bottom of Fig. 6 shows that the beam collected from the objective lens (EFL-2) is designed to pass through relay lens-1 (EFL-3) and relay lens-2 (EFL-4) to form a beam waist at the pinhole plane. At this time, beams within μ pass through the pinhole and are transmitted to the tube lens (EFL-5), while beams exceeding μ are geometrically blocked by the pinhole opening. The hatched area on the right conceptually illustrates that angle selection is maintained even after passing through the pinhole, thereby suppressing background scattering and path deviation components at the sensor surface. This angle selection structurally filters out angular displacement noise caused by scattering, parallelism error (collective tilt of the beam), roughness / local tilt, etc., thereby improving the signal-to-noise ratio (SNR) and contrast.

[0100] Referring to Fig. 6, μ is a sensitivity control parameter, and the critical angle tailored to process requirements can be numerically set by exchanging / variing h or adjusting feq. The alignment criterion in the pinhole plane is optical axis matching, allowing for initial alignment and period correction so that the energy center of the normal path passes through the pinhole center. Additionally, since this structure filters out angular displacement via μ even in non-planar specimens, it can reduce z-scan dependency and increase throughput.

[0101] Meanwhile, although the relay lens group can be selectively changed according to spatial constraints or magnification requirements, it is clearly evident that the key is the position of the pinhole (Poirier / conjugate plane) and the setting of μ. In short, Figure 6 structurally illustrates the operating principle of the optical inspection system of the present invention, in which the pinhole aperture h and the receiving half-angle μ block defect-inducing angular displacement components and acquire only the normal path components with high contrast through the imaging path leading from the relay to the tube lens to the sensor. Due to this principle, the effects of background diffuse reflection, parallelism error, and roughness-induced scattering are reduced, and consequently, false positives and missing positives are simultaneously reduced, and stable high-precision inspection becomes possible even for curved and irregular specimens.

[0102] FIG. 7 is a conceptual diagram illustrating the change in the light path when light irradiated from a light source travels along a normal path and when it does not, in the optical system configuration according to the present invention.

[0103] During normal path progression, light emitted from the light source is focused as it sequentially passes through the objective lens (EFL-2), relay lens-1 (EFL-3), and relay lens-2 (EFL-4). At this time, light rays located within an angular range of μ or μ2 with respect to the optical axis can pass through the DOT-Aperture (pinhole). The passed light passes through the tube lens (EFL-5) and reaches the image sensor to form an image.

[0104] When light travels outside the normal range, if it travels at an angle exceeding μ or μ2 with respect to the optical axis due to factors such as refraction, reflection, scattering, or diffraction, the ray is blocked at the DOT-Aperture. This is intended to suppress path deviations caused by surface defects, tilt, height differences, etc., as shown in the example of optical path deviation.

[0105] That is, light irradiated from a light source reaches the DOT-Aperture via the Objective Lens (EFL-2), Relay Lens-1 (EFL-3), and Relay Lens-2 (EFL-4) along a designed normal path as shown in FIG. 7. The light ray corresponding to the normal path is located within an angular range of μ or μ2, and this light ray passes through the aperture (pinhole) of the DOT-Aperture, passes through the Tube Lens (EFL-5), and is formed on the image sensor.

[0106] Conversely, if a light beam deviates from its normal path due to the influence of defects on the measurement target surface, local tilt, refraction, reflection, scattering, or diffraction, it exceeds the angle limit of the DOT-Aperture. Such light beams are blocked at the pinhole and do not reach the image sensor, thereby effectively eliminating unnecessary scattered, reflected, and ambient light.

[0107] This configuration enables imaging and analysis of cases where light deviates from its path, as in the example of optical path deviation, while simultaneously providing the effect of enhancing the contrast of the ray along the normal path. In particular, by allowing only rays incident at an angle of μ or less to pass through, the resolution and signal-to-noise ratio (SNR) of the imaging result can be increased, and the precision of surface shape measurement can be improved.

[0108] In Fig. 7, h2 and h3 represent the height (radius) formed by the optical axis (reference axis) and the ray at each lens position, and μ and μ2 represent the permissible angle range at the front and back of each lens.

[0109] To explain the relationship between μ and h2 in the section between the objective lens and Relay Lens-1, when light from a light source passes through the objective lens, only light within the angle-accepting range μ is transmitted through the lens along the normal path. Here, h2 represents the radius between the optical axis and the ray at the plane of incidence of Relay Lens-1, and is determined by the product of μ and the lens focal length. Rays outside the μ range are subsequently blocked by the lens or become out of focus.

[0110] Looking at the relationship between μ2 and h3 in the interval between relay lens-1 and relay lens-2, the light passing through relay lens-1 has a new angular range called μ2. μ2 can be set according to the combination of lens arrangement and focal length between relay lens-1 and relay lens-2. In this interval, h3 is the radius from the exit plane of relay lens-2 to the optical axis and is determined by μ2 and the lens design.

[0111] Light passing through relay lens-2 is collected by the DOT-Aperture, and h = h3 corresponds to the radius of this pinhole. The size of the pinhole is directly linked to the final angle acceptance range allowed by the system, and light corresponding to a radius of h3 or less (angles of μ2 or less) passes through the pinhole and reaches the sensor, while light corresponding to a radius greater than h3 (angles greater than μ2) is blocked at the pinhole.

[0112] Consequently, in the normal path, if light irradiated from the light source travels within the range μ → μ2, it maintains the path h2 → h3 → h3 (pinhole radius), passes through the pinhole, and forms an image on the sensor. If the light deviates from the normal path, light exceeding the μ or μ2 range due to refraction, reflection, scattering, or tilt (θ) deviates too far from the optical axis at h2 or h3 and is blocked at the pinhole.

[0113] Through this structure, an angular displacement suppression effect as shown in Figs. 2 and 3 can be realized.

[0114] Figure 8 explains the definition and design criteria of a pinhole aperture. In Figure 8, a circular pinhole aperture is positioned in the center of a black disc, and the path of light passing through the pinhole is shown as a yellow line.

[0115] The pinhole aperture used in the present invention is installed at a position where the light of pure illumination is focused and may be formed as a thin glass plate or a central disc structure having a circular opening of a certain size. The pinhole aperture is fixed by a support to such an extent that it does not obstruct the maximum passage of light, and is designed to allow only light within an angle range specified in the optical system to pass through.

[0116] The size of the pinhole aperture is set based on the boundary between the normal optical path and the abnormal optical path. That is, light incident at an angle μ or less is allowed to pass through the pinhole opening, while light incident at an angle exceeding μ is blocked. The height (h3) of light incident at an angle μ at the Dot-Aperture position corresponds to the radius of the pinhole and is defined by the following mathematical formula.

[0117]

[0118] Here f 2 : Focal length of the objective lens, f 3 : Focal length of relay lens-1, f4 : Focal length of relay lens-2, μ represents the maximum allowable angle of incidence.

[0119] The pinhole aperture designed in this way effectively blocks optical path deviation components caused by refraction, reflection, scattering, etc., by ensuring that only light incident at an angle of μ or less reaches the sensor. As a result, it provides the effect of improving image contrast and sharpness.

[0120] A pinhole is positioned at a location where pure illumination is focused, and is formed as a circle of a certain size, a dot on a thin glass plate, or a central disc supported by a support that minimizes obstruction of light passage.

[0121] The size of the pinhole is set according to the boundary separating the normal path and the abnormal path, where the boundary is defined by an angle μ representing the change in the light propagation path.

[0122] The height at which light incident at an angle μ reaches the dot-aperture position becomes the criterion for determining the size of the pinhole.

[0123] FIG. 9 is a diagram illustrating the entire optical path in which light from an object passes through multiple lenses and apertures to reach the image plane in an optical system. First, light emitted or reflected from the object, indicated by the black square on the far left of the diagram, passes through a first lens group, where it is focused and its path is adjusted. Subsequently, at an aperture or pinhole located in the middle of the optical path, only light within the μ angle range is allowed to pass through, while other light rays are blocked. As an Aperture Stop at the intermediate point, the aperture is located in the middle of the optical path and serves to reduce system aberrations and increase contrast by blocking unnecessary light rays. The light, having been angle-limited in this way, passes through a rear lens group where it is refocused to minimize distortion and reaches the image plane in a focused state according to the Abbe compensation design. Finally, an image of the object is accurately formed on the image plane, which can then be recorded on a sensor or film.

[0124] The Aperture Stop at the midpoint refers to the aperture and serves as a point that limits the maximum range of light rays that can pass through the entire optical system. Its function is to control exposure and brightness by limiting the amount of light passing through, and to reduce aberrations (especially spherical and coma aberrations) and scattered light by controlling the maximum angle of incidence of the light rays. Additionally, it controls the depth of field by increasing the size of the Aperture Stop, which increases the depth of focus, and decreasing it, which decreases it. Its location is positioned at the center of the optical system or between specific lens groups to optimize the angle-limiting effect across the entire light path. In the case of FIG. 9, it is positioned between the front and rear lens groups to block light exceeding the μ angle.

[0125] In contrast, the pinhole is positioned just before image formation and is configured to selectively pass only light whose path has not been altered (maintaining the regular path). Unlike an aperture, its purpose is not merely to limit light intensity, but rather to focus on suppressing optical path displacement and blocking scattered light. Based on the angle μ, it ensures high-precision imaging by allowing only light within the allowable angle to pass through while blocking the rest.

[0126] In addition, the present invention provides an optical inspection system capable of significantly improving the accuracy of defect detection in curved or non-curved specimens. The system is composed of a single light source, a multi-stage lens group (Objective, Relay, Tube Lens), a pinhole, and a sensor, and can maintain focus on non-curved specimens and obtain accurate inspection results by adjusting the F / # value.

[0127] The optical inspection system of the present invention eliminates light rays exceeding the allowable half-angle μ range and allows only normal path light rays to contribute to imaging, thereby enabling high-contrast images to be secured even for specimens with complex shapes or curvatures.

[0128] The optical inspection method of the present invention provides homogeneous focus and high contrast across the entire area of ​​a non-flat specimen through the selection of light rays based on the allowable half-angle μ standard, control of the angle of incidence by the aperture, and expansion of the depth of field as F / # increases.

[0129] Importantly, in describing the position of apertures such as point apertures or pinholes in this specification, the respective acceptance range (μ, etc.) and the height (radius) (h, etc.) formed by the optical axis and the light ray at the lens position are described with the optical axis as the reference axis; however, this is merely an illustrative description. The reference axis may be set in various ways depending on the design requirements or configuration of the actual system, and therefore, the scope of the present invention should not be interpreted as being limited to the case where the optical axis is the reference axis.

[0130] The attached drawings are not drawn at full scale but are partially enlarged and reduced to explain the technical concept of the present invention.

[0131] Furthermore, it is obvious that embodiments in which the components and steps of the system, method, or device of the present invention described above are selectively combined also fall within the scope of the present invention.

Claims

Claim 1 An optical inspection system for inspecting surface defects of a non-flat specimen, wherein the optical inspection system comprises: an illumination unit that irradiates the specimen with a single light source; an object lens that receives light reflected or refracted from the specimen to form a primary image; a relay lens that adjusts the optical path as necessary; a tube lens that forms a final image; a pinhole that allows only light within an allowable angle (μ) range to pass through; and a sensor that receives light passing through the pinhole to form an image. The pinhole is positioned on a polyhedral plane or a conjugate plane where a beam waist is formed by passing through the relay lens. The entire specimen is positioned within the focal range by expanding the depth of field through an increase in the F / # value of the optical inspection system. The pinhole allows only normal light rays to pass through and blocks light rays whose paths have been displaced by reflection, refraction, or shielding. It distinguishes between normal and abnormal light rays based on the allowable angle μ at the pinhole position and blocks light rays exceeding μ, thereby enabling simultaneous achievement of focus maintenance and angle selection in the non-flat specimen. System. Claim 2 An optical inspection system according to claim 1, characterized in that the single light source provides a diffused beam. Claim 3 delete Claim 4 An optical inspection system according to claim 1, characterized by additionally placing an aperture stop on the optical path to limit the maximum aperture and angle of incidence of the incident light. Claim 5 An optical inspection system according to claim 1, characterized by selectively passing only light rays within a μ range based on the local height difference (h) and local inclination (θ) of the specimen relative to the reference plane. Claim 6 An optical inspection system according to claim 1, characterized in that the size of the pinhole is determined by setting the boundary between a normal ray and an abnormal ray to an angle μ. Claim 7 An optical inspection method using an optical inspection system for inspecting surface defects of a non-flat specimen, wherein the optical inspection method comprises the steps of: illuminating a specimen with a single light source; receiving light reflected or refracted from the specimen with an object lens to form a primary image; adjusting the optical path with a relay lens if necessary; performing a final image with a tube lens; allowing only light within an allowable angle (μ) range to pass through using a pinhole; and receiving light that has passed through the pinhole with a sensor to form an image, wherein the pinhole is positioned on a polyhedral plane or a conjugate plane where a beam waist is formed by passing through the relay lens, and the F / # value of the optical inspection system is increased to expand the depth of field, wherein the pinhole allows only normal light rays to pass through and blocks light rays whose paths have been displaced by reflection, refraction, or shielding, and wherein normal light rays and abnormal light rays are distinguished based on the allowable angle μ at the pinhole location, and light rays exceeding μ are blocked, thereby simultaneously achieving focus maintenance and angle selection in the non-flat specimen. Claim 8 An optical inspection method according to claim 7, characterized in that the single light source provides a diffused beam. Claim 9 delete Claim 10 An optical inspection method according to claim 7, further comprising the step of additionally placing an aperture on the optical path to limit the maximum aperture and angle of incidence of the incident light. Claim 11 An optical inspection method according to claim 7, comprising the step of selectively passing only light rays within a μ range based on the local height difference and local inclination of the specimen relative to the reference plane. Claim 12 An optical inspection method according to claim 7, characterized in that the size of the pinhole is determined by setting the boundary between the normal light ray and the abnormal light ray to an angle μ.

Images