System and method for inspecting moist contact lenses

Abstract

Observation system for inspecting a contact lens (18) being transported in an automated manufacturing plant, which includes the following: a plurality of lighting modules (14, 15, 16, ..., N2) for illuminating the contact lens (18), which emit light, comprising various lighting techniques; a plurality of beam splitters (11, 12, 13, ..., N4) for directing the plurality of illumination to highlight the contact lens (18); a plurality of beam splitters (8, 9, 10, .... N3) for receiving a plurality of processed images of the contact lens; a plurality of filters (5, 6) for filtering the images by a suitable method; a plurality of optical modules (5, 6, 7, ..., N2) for obtaining a plurality of images of the contact lens (18), wherein each of the plurality of optical elements processes the image to enhance the contrast of different defects of the contact lens (18); a plurality of optical detectors for receiving filtered and processed images; a contact lens holder (17) for transporting the contact lens; an image analysis device for processing and analyzing the majority of images for defects in the contact lens, wherein said image analysis device is configured to extract an edge from the majority of the images; wherein said image analysis device is further configured to draw two or more concentric circles on the analyzed majority of the images and to transform an area of ​​interest encompassed by the two or more concentric circles to produce a two-dimensional image; and wherein said image analysis device is operable to analyze an inverted contact lens (18) by defining an area in the two-dimensional image by a small box, binarizing the area by defining a light intensity limit for dark and white pixels, identifying a group of white pixels in the area and evaluating the geometric property of at least one group of white pixels.

Description

Technical area:

[0001] The present invention relates to a system and a method for inspecting moist contact lenses, preferably in an automated production line. The inspection unit provides a system of multiple cameras with high-resolution imaging sensors, configured to enable image acquisition using multiple inspection channels. Each channel consists of at least one camera with a customized optical module and an illumination module for highlighting the contact lens to be inspected. The optical module for each channel can further include beam splitters and associated lens components to process the light beams to produce dark-field or bright-field images, depending on which defects are to be inspected.The present invention also provides a method for inspecting the aforementioned images through the aforementioned channel, which is set up for certain defect types, and which significantly increases the inspection quality. Background of the invention:

[0002] Contact lenses are widely used, and the high demand has necessitated the large-scale production of both standard and cosmetic lenses to a very high standard. It is a well-known fact that lenses manufactured in automated production facilities are more reliable than those produced using manual systems, which are prone to unforeseen problems. It is also widely recognized that monitoring systems are an essential component of automated lens inspection facilities, ensuring a consistently high-quality inspection process and delivering premium products to customers.The regular improvement of parameters and the generation of configuration files containing inspection characteristics for various contact lens models allows for flexible adaptation of the inspection system to different types of lenses. Contact lenses are intended for use in the human eye not only to correct vision but also to enhance the cosmetic appearance of the eye through the printing of designs onto the lens. Therefore, great care must be taken to ensure they are free of defects. These contact lenses are manufactured in large quantities in an automated production facility. To ensure that each lens is produced according to stringent quality control standards, it is essential that the lenses are inspected using automated inspection procedures shortly before packaging.

[0003] Before packaging, the contact lenses are placed in transparent lens holders. Each holder holds one lens, which is typically immersed in a liquid solution. The moist lens in the holder is inspected as the lens carrier moves along a conveyor belt in an automated production line. To increase the production line's throughput, it is essential that the lenses are inspected as quickly as possible.

[0004] The holder used should preferably be made of clear glass without any coating on the underside. If the underside of the holder is coated, this can lead to images with uneven backgrounds.

[0005] The techniques commonly used for inspecting wet contact lenses are bright-field imaging, dark-field imaging, and images acquired under infrared illumination. Various types of illumination are used, including halogen lamps, xenon lamps, and LEDs. Different illumination methods can accentuate different types of defects, making it easy to detect even very small defects and thus minimizing the risk of incorrect rejection.

[0006] In the prior art, DE 693 33 492 T2 discloses a video inspection system with multispectral illumination using light-emitting diodes and a method for multispectral illumination using light-emitting diodes, EP 1 474 648 B1 discloses systems and methods for examining contact lenses, EP 1 224 448 B1 discloses a method and an apparatus for measuring the geometric structure of an optical component by light transmission, and DE 690 21 753 T2 discloses a method for checking deviations of an object or a series of successively supplied objects.

[0007] Prior art US patent 6,765,661 B2 discloses the use of combined bright-field and dark-field imaging to provide high-quality defect inspection. However, it is evident from later patents by the same inventor (prior art US patents 7,663,742 B2 and 7,855,782 B2) that the use of bright-field imaging in combination with dark-field imaging alone is insufficient to comprehensively inspect all types of defects in a contact lens. Prior art US patents US 7,855,782 B2 and US 7,663,742 B2 disclose the use of combining phase contrast imaging with either brightfield imaging or darkfield imaging to inspect all types of defects such as accuracy of size, surface defects, tears, peripheral cracks, inclusions such as bubbles and foreign bodies, and small defects at the edges of contact lenses.In these prior art inventions, a monochromatic light source is used together with beam splitters and other optical elements to divide the light into side beams. One of the side beams is used for phase contrast imaging, while the other is used for either bright-field or dark-field imaging. A complete inspection of the contact lens is achieved by sequentially switching the light source to either bright-field or dark-field imaging in conjunction with phase contrast imaging. This is time-consuming and susceptible to parasitic effects between the illumination and external light. Furthermore, the use of a holder with a flat bottom surface makes it difficult to repeatedly locate the edge of the object being inspected, as it can move around on the holder.The lack of depth of field also affects the sharpness of the image in some areas of the contact lens, which impacts the ability to detect edges and subsequently inspect defects.

[0008] There is therefore a need to capture multiple high-resolution images of the contact lens using several cameras for inspection under different lighting conditions, wavelengths, and polarizations, and to be able to trigger the lighting and camera shutters at the user's discretion, either simultaneously or at different times, without sacrificing image quality. This is the objective of the present invention. Summary of the present invention:

[0009] The present invention solves the problem by providing a device in the form of a system architecture that is suitable for adding or removing inspection stations without taking up too much space, and avoiding the need to move the object from one inspection station to the next, thereby significantly reducing inspection time and increasing the throughput of the manufacturing plant.

[0010] In one aspect of the invention, a device for capturing multiple images of a contact lens is provided using several monochrome cameras, each equipped with an individual optical module that uses or omits filters for different wavelengths. The object to be inspected (the contact lens) is illuminated by several illumination modules, each configured for a different wavelength or polarization and capable of being triggered to flash the illumination at different times or simultaneously within the same time zone.

[0011] In a further aspect of the invention, the present invention provides a method for capturing multiple images of a contact lens using several monochrome cameras, each equipped with an individually adapted optical module that uses or does not use different wavelength filters. The object to be inspected (the contact lens) is illuminated by several illumination modules, each configured for a different wavelength or polarization and triggered to flash the illumination at different times or at the same time within the same time zone.

[0012] In another aspect of the invention, the system allows the simultaneous acquisition of multiple darkfield, brightfield and high-contrast images of the object or differently illuminated images of the object by cameras of multiple inspection stations.

[0013] In another aspect of the invention, the system allows the separate acquisition of multiple darkfield, brightfield and high-contrast images of the object or differently illuminated images of the object by cameras from several inspection stations through the selective flashing of the respective required illumination module at different times.

[0014] In yet another aspect of the invention, the present invention provides a method for taking and inspecting multiple images of the object, which is held in various forms of microscope slides.

[0015] In another aspect, the invention provides a device and a method for recording and inspecting multiple images of the object, which are printed, colored or simply transparent contact lenses.

[0016] Further details and advantages of the invention can be derived from the following description and the illustrations. Brief description of the illustrations: It will be helpful to further describe the present invention with reference to the attached figures, which illustrate possible arrangements of the invention. Other arrangements of the invention are possible, and accordingly, the specificity of the attached figures should not be understood as negating the generality of the preceding description of the invention. Fig. Figure 1 shows an inspection system according to the present invention with multiple lighting modules and multiple cameras. Fig. Figure 2 shows a different lens holder for dry lens inspection, usually after the forming process. Fig. Figure 3 shows another holder with a flat surface on the underside, which usually contains a liquid (not shown). Fig. Figure 4 shows a graph of the spectral sensitivity of typical illumination devices of wavelengths 1, 2 & 3 (WL1, WL2 and WL3) of the present invention. Fig. Figure 5 shows a table of the typical defects that are usually inspected in the three different illumination channels by three different wavelengths of illumination (for example, the illumination modules emit light with wavelengths of 650 nm, 550 nm & 450 nm). Fig. Figure 6 shows the inspection flow diagram for the system of the present invention. Fig. Figure 7 shows the edge of a normal lens, which is located in step 102 of the inspection flow diagram. Fig. Figure 8 shows the unfolded image of part A of the lens image from step 102. Fig. Figure 9 shows an illustration of a small part taken from the unfolded part of the Fig. 8 was extracted. Fig. Figure 10 shows the illustration of the Fig. 9 after binaryization. Fig. Figure 11 shows the edge of the inverted lens, which was located in step 102 of the inspection flow diagram. Fig. Figure 12 shows the unfolded image of part B of the lens image of step 102. Fig. Figure 13 shows the illustration of a small part that results from the unfolded illustration in Fig. 12 was extracted. Fig. Figure 14 shows the illustration of the Fig. 13 after binary conversion. Fig. Figure 15 shows a dark-field image of a normal lens. Fig. Figure 16 shows an enlarged image of part C of the dark-field image of the Fig. 15. Fig. Figure 17 shows a bright-field image of a normal lens. Fig. Figure 18 shows an enlarged image of part E of the brightfield image of the Fig. 17. Fig. 19 is the image that results from the superimposition of the images in Fig. 16 and Fig. 18 results. Fig. Figure 20 shows a dark-field image of an inverted lens. Fig. 21 shows an enlarged image of part D of the dark-field image of the Fig. 20. Fig. Figure 22 shows a brightfield image of the inverted lens. Fig. Figure 23 shows an enlarged image of part F of the bright-field image of the Fig. 22. Fig. 24 is the image that results from the superimposition of the images in Fig. 21 and Fig. 23. Fig. Figure 25 shows the brightfield image of a lens with tear defects. Fig. Figure 26 shows an unfolded image of the edge region of the lens. Fig. 25. Fig. Figure 27 shows an image of a lens with a gap defect. Fig. Figure 28 shows an unfolded image of the edge region of the lens. Fig. 27. Fig. Figure 29 shows an enlarged image of the area where the gap defect is present. Fig. Figure 30 shows an image of a lens with a double edge defect. Fig. Figure 31 shows an unfolded image of the edge region of the lens. Fig. 30. Fig. Figure 32 shows an enlarged image of the area where the double-edge defect is present. Fig. Figure 33 shows an image of a lens with a non-circular defect. Fig. Figure 34 shows an unfolded image of the edge region of the lens. Fig. 33. Fig. 35 and Fig. Figure 36 each shows a dark-field and a bright-field image of a lens with bubble defects. Fig. 37 shows a table illustrating the representation at different limit values ​​for the range X1 and the range X2, which are in the Fig. 35 and Fig. 36 are shown, reproduced. Fig. Figure 38 shows a distribution curve of white to black pixel ratios on the X-axis and the number of samples on the Y-axis. Detailed description of the invention:

[0017] According to the present invention, a preferred embodiment is described with reference to the aforementioned illustrations.

[0018] The system of the present invention uses multiple cameras and optical modules to capture high-resolution images of the object, which is illuminated by several lighting modules. The lighting modules have different wavelengths and can also have different polarizations. By using different illumination wavelengths, the system is resistant to parasitic effects between the lighting modules during image acquisition, particularly when they are flashed or triggered simultaneously during image acquisition. The system can also include color filters to exclude light wavelengths other than the filter wavelength. The system's prevention of interference between the captured images produces images that exhibit amplified defects.This results in quality images that significantly improve the quality of the inspection in terms of both accuracy and inspection speed.

[0019] Furthermore, it has been found that depth of field is a frequently compromised feature in prior art systems. The end result of images captured with systems using shallower depths of field, as found in prior art, is a lack of clarity for small defects located in areas of the captured image that fall outside the depth of field. Depending on the optical adjustments made for specific product types, images of contact lenses may appear sharp in certain regions and blurred in others. Inaccurate defect measurement leads to defective lenses entering the market. Additionally, achieving greater depth of field in multi-camera systems has been found to be a significant challenge.

[0020] The optical system of the present invention is designed to have a depth of field that corresponds almost exactly to the curvature of the lens. The curved depth of field facilitates the acquisition of sharp images of the entire lens profile. Locating the object in the holder is faster because the entire lens is well focused, regardless of whether the holder has a curved or a flat recess for holding the lens. The resulting image, when visualized by a user or analyzed by a computer program, makes tears, blisters, cuts, and defects with contrast easily visible and detectable.

[0021] As in the Fig. As shown in Figure 1, the system of the present invention uses “N” illumination modules and “N” cameras, a specially designed holder, beam splitters for directing the light from the illumination module to the contact lens and then to the corresponding camera through specially designed optical modules for each camera, which depend on the type of feature to be inspected.

[0022] In a preferred embodiment of the invention, the illumination modules are flashed simultaneously to obtain multiple images at once. Suitable wavelength filters are used to eliminate any parasitic effects. The system uses monochrome cameras to achieve higher-resolution images. Furthermore, the different illumination modules can be flashed at different time intervals, and the various cameras are synchronized with the illumination flash pulses to capture images. In some cases, there is a need to introduce time delays before image capture to meet the intensity requirements of the image. Delays are typically incorporated for the camera shutter release when the time required for the illumination to reach maximum intensity is slower.For example, depending on the timing of the shutter release, images tend to become blurry or faint if the cameras are triggered at the same time as the illumination. To enable an image of uniform intensity, it is preferred that the camera shutter be delayed by at least 50 ms to allow the light intensity to reach its saturation point, after which the camera shutter is released. This technique results in uniform images, although the consequence of this method is that the illumination flash pulse is typically longer than the camera shutter pulse. The illumination control mechanism is not shown, as it is outside the scope of this invention.

[0023] The optics of each of the "N2" channels are designed differently to suit the varying illumination characteristics of the object. The illumination system utilizes illumination modules of different wavelengths to produce dark-field or bright-field images. The choice of illumination can vary depending on the type of defect. In such cases, the number of cameras and their associated individually adapted optics and illumination systems can be increased to target specific defect types that are best visualized at certain wavelengths and magnifications. Furthermore, the image acquisition method, whether performed independently or simultaneously, can be evaluated to achieve optimal imaging of the specific defect.

[0024] In an alternative embodiment, the monochrome cameras can be replaced by color cameras to obtain color images at a lower resolution. The captured color images are suitable for extraction in three different colors, which can be further analyzed using appropriate algorithms to detect defects in the extracted images.

[0025] It should be mentioned that the present invention relates to an imaging system that uses multiple imaging devices and illuminations to enable highly accurate and sharply focused images of the object (the contact lens). Only then is said imaging suitable for the efficient detection of minute features or defects in the contact lens, including but not limited to tears, cuts, bubbles, inclusions, cracks, deformations, size errors, and foreign material contamination.

[0026] The system of the present invention allows the acquisition of multiple brightfield, darkfield, and high-contrast images. The contact lens to be inspected is held in a glass holder and suspended in a liquid. The holder is preferably made of clear glass without any coating on the underside to prevent any inconsistency in the captured image. Non-uniform images are obtained when a glass holder with a coating on the base is used. For this reason, it is recommended that a clear glass holder without any coatings on the underside be used. Regardless of the design of the holder in which the contact lens is suspended in the liquid, the optical system is configured to have a depth of field that closely corresponds to the curvature of the contact lens.For example, if the lens holder has an inner radius of 14 mm and the lens has a radius of 8 mm, the optical system is designed to have a curved depth of field with a radius of 11 mm, ensuring that the entire lens is in focus. This curved depth of field allows the lens to be located even if it is slightly off-center.

[0027] The holder, in which the contact lens is positioned, contains a liquid such as water, saline solution, or a similar transparent solution. The holder is typically designed so that the contact lens is automatically centered at its lower end. However, different holder types can be used to hold the lens, but due to the curved profile of the depth of field built into the optical module, the entire lens is in focus. The holder can be used as a standalone device or can be part of a larger mechanism that includes multiple holders.

[0028] In another embodiment of the invention, the holder holds the lens upside down as in the Fig. Figure 2 shows lenses positioned on such holders. These are commonly found in molds used to manufacture the lenses. In another embodiment of the invention, a standard holder is used to position the lens under the inspection system. The holder has a flat bottom surface, as shown in Figure 2. Fig. 3 shown, and the lens is suspended inside the holder in a liquid (not shown). Detailed description of the illustrations:

[0029] With reference to Fig. The inspection system consists of several cameras 1, 2, 3 up to a number N, which are mounted together with optical modules 5, 6, 7 up to a number N2. The cameras are preferably monochrome cameras to capture high-resolution images. Color cameras can also be used instead of the 5, 6, 7 and N sensors, but the image resolution is lower compared to monochrome cameras of the same resolution.

[0030] Those skilled in the art will recognize that both monochrome and color cameras can be used in the system to serve the purpose of the invention. The arrangement of the optical modules 5, 6, 7 up to N2 varies depending on the type of defects to be recorded and the type of illumination to be used.

[0031] The lighting modules 14, 15, 16 up to N5 are configured for different light wavelengths WL1, WL2, WL3 etc. to cover the different requirements listed in the table. Fig. to highlight the defects shown. Additional such illumination modules can be combined with suitable optical modules to enable the detection of new types of defects. In principle, the inspection system can be configured as required by adding new inspection channels. Beam splitters 11, 12, 13 up to N4 direct the light of a specific wavelength, which emerges from the illumination modules 14, 15, or 16 up to N5, to the contact lenses 18 suspended in the holder 17. Beam splitters 8, 9, 10 up to N3 direct the light passed through the lens 18 under inspection to the individually adapted optical modules 5, 6, 7 up to N2. Finally, images are captured by cameras 1, 2, 3 up to N after being processed by the optical modules 5, 6, 7 up to N2.In fact, N images with different characteristics are taken for the inspection of various features in the contact lenses.

[0032] The previously described holder 17 consists of monochromatic lenses and has no coating on its surface to minimize image distortion.

[0033] In the preferred embodiment, the holder 17, which contains the contact lens 18 immersed in a liquid (not shown), has a curved bottom surface to conform to the profile of the inspected contact lens.

[0034] In another embodiment as in the Fig. The holder shown can be the inverted type 19, which is suitable for inspecting the lens in its dry state immediately after the forming process.

[0035] In another embodiment as in the Fig. As shown in Figure 3, the holder 20 can have a surface on the underside that is of the flat type, which is also suitable for the inspection of moist lenses.

[0036] The Fig. Figure 4 shows the spectral sensitivity of the three illumination modules, which provide light with three different wavelengths. WL1 has a range of 400 nm to 500 nm. Typically, WL1 is set up for 430 nm. WL2 has a range of 500 nm to 600 nm. Typically, WL2 is set up for 550 nm. WL3 has a range of 600 nm to 700 nm. Typically, WL3 is set up for 650 nm.

[0037] The light exiting the illumination module 14 has a wavelength WL1, which ranges from 400 nm to 500 nm. Typically, WL1 is set to 450 nm. This light comprises scattered and reflected light rays, which produce a dark-field image. Dark-field images highlight defects with very low contrast and, in certain cases, also enhance defects without contrast. Light from the illumination module 14 is deflected by the beam splitter 11 onto the optical lens 18, which is immersed in liquid within the holder 17. The beam splitter 8 directs the light exiting the optical lens onto the optical filter 5 of the imaging system, which includes the camera 1. Column 4 of the table in the Fig. summarizes the defects that are covered by the first inspection channel, which uses light of wavelength WL1.

[0038] The light exiting the illumination module 15 has a wavelength WL2, which covers a range from 500 nm to 600 nm. Typically, WL2 is set to 550 nm and highlights defects to produce high-contrast images. The camera 2 is set with a small aperture to capture images resulting from the illumination provided by the light head 15 at wavelength WL2. At smaller apertures, the illumination of the optical lens is at a very narrow angle, which helps to capture images with good contrast for most defects. Light from the illumination module 15 is directed by the beam splitter 12 onto the contact lens, which is immersed in liquid within the holder 17. The beam splitter 9 directs the light exiting the contact lens to the optical filter 6 of the imaging system, which comprises the camera 2. Column 3 of the table in Fig. Section 5 summarizes the defects that are covered by the second inspection channel, which uses the wavelength WL2.

[0039] The illumination module 16 can be configured for wavelength WL3, which has a range of 600 nm to 700 nm. Typically, WL3 is configured for 650 nm. Light of this wavelength highlights defects to produce brightfield images, while the optical system guides the beams through the contact lens, producing either parallel, diverging, or converging beams. The light from the illumination module 16 is directed by the beam splitter 13 onto the contact lens 18, which may be immersed in liquid in the holder 17. The beam splitter 10 directs the light exiting the contact lens onto the optical module 7 of the imaging system and onto the camera 3. Light at wavelength WL3 has a wide beam and produces an image with uniform brightness across the field of view. Brightfield imaging, which uses this light, produces uniform images over a large field of view.Defects across the entire lens are easily measured because the illumination intensity is uniform. It should be noted that the image produced by light of wavelength WL3 is not sensitive to the position of the contact lens in the liquid, and a good image of the contact lens can be captured regardless of its displacement. Column 2 of the table in . Fig. summarizes the defects that can be detected by the third inspection channel, which uses light of wavelength WL3.

[0040] An expert will recognize that the evaluation can be carried out using different wavelengths of light and that a table similar to the one in the Fig. The results shown can be obtained. Based on the study, suitable lighting modules, optical modules, wavelength filters, and cameras can be selected to design the inspection channel. The system's architecture allows for the easy addition or removal of inspection channels.

[0041] Examples of preferred illumination modules include light-emitting diodes or short-arc xenon flash lamps. Other illumination modules, such as halogen lamps, can be used, although suitable filters may be required to achieve optimal image quality.

[0042] In a further embodiment of the invention, an inspection channel can be provided to capture images with illumination modules that operate in the infrared spectrum.

[0043] In a further embodiment of the invention, an additional inspection channel can be configured to examine features such as print quality.

[0044] In the preferred embodiment of the invention, all lighting modules are flashed at the same time and all cameras simultaneously capture different images according to the respective light setting.

[0045] In an alternative embodiment of the invention, the lighting modules are flashed at different times and the associated cameras capture images according to the light setting.

[0046] In yet another embodiment of the present invention, the lighting modules used can be selectively switched off depending on the type of defects to be examined.

[0047] The lighting control (not shown) is CPU-controlled to adjust the flash intensity and pulse duration. The CPU also controls the timing of the trigger pulse synchronization with the camera aperture to maintain consistent image quality. For clarity, the flash mechanism and image capture technique are not discussed, as transferring images to computer storage is a well-established technology. Images are then moved or copied to different storage locations for further image processing.

[0048] We now refer to Fig. 6. It shows the flowchart for the inspection procedure. Software algorithms begin processing (in step 1) the stored images, starting with step 100. In this step, the lens mount is detected first, followed by the lens edge in step 101. If the software detects the lens within the mount, the inspection sequence moves to step 102. If the lens edge cannot be found, the sequence moves to step 103 to display an error message and ends in step 120. In step 102, the lens is located using edge-finding algorithms, after which a decision is made regarding the presence or absence of a lens. In step 102, the lens edge is then located and its position is recorded. Fig. 7. The lens edge is indicated by an outer dashed circle line.

[0049] A first method is based on the lens structure and its properties. In this method, in step 104, a lens as described in Fig. The concentric area of ​​the lens shown from the lens edge is selected and unfolded for further processing. A second concentric circle is drawn relative to the outer circle at a programmable distance from the center of the lens. The area inside the two concentric circles is then processed as shown in Fig. Figure 8 is shown unfolded. It should be noted that the unfolded image is not to scale. Many edge defects are detected by using the unfolded portion of the image.

[0050] In the next step, 105, an inverted lens is inspected. A small area A is selected as programmed by the user. For example, area A is extracted ( Fig. 9) and can be processed using a binaryization technique. The one in the Fig. The resulting image shown in Figure 10 depicts an image with white vertical lines.

[0051] Another in Fig. Image 11 is shown in the same way as in Fig. 7 is processed by selecting and binaryizing another area, such as B. The resulting binaryized image in Fig. 14 shows horizontal lines.

[0052] The lens structure is such that a normal lens, due to its poor signal-to-noise ratio characteristics, exhibits vertical lines in its binarized image, while an inverted lens exhibits horizontal lines. Based on the orientation of the lines in the binarized image, a decision is made in step 105. If the lens is determined to be inverted, the process proceeds to step 109. If the lens is found to be normal or not inverted, the process proceeds to step 106, where the next set of defects is detected.

[0053] A second method for detecting an inverted lens is described here. Fig. Figure 15 shows a dark-field image of a normal lens. Fig. Figure 16 shows an enlarged image of part C of the lens in Fig. 15. The edge of the lens shows a white, curved, thick line. Fig. Figure 17 shows a brightfield image of the same lens. Fig. Figure 18 shows an enlarged image of part E of the bright-field image in Fig. 17. Obviously, the dark edge of the bright-field image is thicker than the white edge of the dark-field image in Fig. 16. If the image in Fig. 18 with the image in Fig. When 16 is superimposed, the resulting image shows in Fig. 19 a thin white edge followed by a dark edge.

[0054] Fig. 20 is a dark-field image of an inverted lens and Fig. Figure 21 is an enlarged image of part D of the lens in Fig. 20. The edge shows a white edge, which corresponds to the image of a normal lens in Fig. 16 is quite similar. Furthermore, Fig. 22 a bright-field image of an inverted lens and Fig. Figure 23 is an enlarged image of part F of the lens in Fig. 22. The dark edge of the bright-field image in Fig. 23 is thinner compared to the dark edge in the bright-field image of a normal lens as in Fig. 18. If the image in Fig. 23 and the image in Fig. The resulting image, when 21 are superimposed, shows in Fig. 24. A white edge followed by a pale dark line. The phenomenon that, in the case of an inverted lens, the dark edge does not extend beyond the white edge is a key feature being investigated to determine the difference between a normal lens and an inverted lens. After examining many different samples of the lens, it was found that a percentage threshold can be obtained by using the width of the white edge and the width of the dark edge. This parameter can be used to accurately detect a normal and an inverted lens.

[0055] If the lens is detected as normal, the inspection proceeds to step 106, where any defects relating to the lens edge are examined.

[0056] Fig. Figure 25 shows an image of a contact lens with a tear defect. In this image, the procedure for locating the lens, the lens edge, and unfolding it is similar to the steps described in Figures 102 and 104 of the [reference missing]. Fig. The flowchart shown in section 6 explains this. After the image is unfolded, the software checks the height of the tears L1 and L2 ( Fig. 26) with regard to the edge of the lens. Based on user-defined defect criteria, the program decides whether the lens tear is a defect or not.

[0057] Fig. Figure 27 shows another image of a contact lens with a gap defect. In this image, the same procedure for edge extraction and unfolding of the concentric area is performed. An unfolded image is shown in Fig. 28 shows an enlarged area of ​​a gap defect. Fig. 29 shown and the position of the defect in the unfolded image is shown in Fig. 28 is displayed. Here again, the program decides whether the lens is defective or not, based on defect criteria defined by the user.

[0058] Fig. Figure 30 shows an image of a double-edged contact lens. The enlarged image Fig. 32 and the unfolded image in Fig. Figure 31 shows a white line between two black lines, indicating a double-edged lens. This phenomenon of a white line between the black lines is called a double-edged defect. By comparing the dimensions of this defect with the defect criteria defined by the user, a decision is made as to whether the lens is defective or not.

[0059] Fig. 33 and Fig. Figure 34 shows an image of a non-circular lens. The procedure for locating the edge and unfolding the lens surface is carried out according to steps 102 and 104. The distance from the line edge to the non-circular surface is measured and then compared with the rejection criteria defined by the user to decide whether the lens should be rejected or not.

[0060] Although some processing methods are explained, a specialist will recognize that the software can be configured to measure other types of edge inclusions. Once the edge defect analysis is complete, the process moves to step 107, where the program checks whether any defects were found in step 106. If so, the process jumps to step 109 and then to step 120.

[0061] If no defect is found in step 107, the process moves to step 108. Here, the procedure for detecting defects within the lens begins. The inspection procedure described below relates to bubbles found within the lens and air bubbles typically found on the lens surface. To reduce the excessive distortion caused by air bubbles, a new method is explained below. A person skilled in the art will recognize that the same arrangement of algorithms, with minor modifications to the procedure steps, can be used to detect most defects within the lens without requiring changes to the underlying algorithm.

[0062] Fig. 35 and Fig. 36 show a dark-field image of a lens and Fig. Figure 36 shows a bright-field image of the same lens as in Fig. 35. After the analysis, two points X1 and X2 were selected. The two areas X1 and X2 were binarized at different limit values, and a total of four images were generated in the Fig.The results are recorded in the table shown in Figure 37. Image 2 of X1 shows a larger white area than the white area of ​​X2 in Image 2. However, Image 1 of X1 shows a smaller, darker area compared to Image 1 of X2. The same applies to Images 3 and 4. After some experimentation, it was found that a bubble is present inside the lens when the ratio of white area to black area is greater than 35%. However, for a ratio of white area to black area of ​​less than 25%, an air bubble is present at this location. For any ratio between 25% and 35%, further processing may be necessary, which may include a superimposition method or other techniques. The superimposition method involves the process of overlaying two images, in this case, the images in columns 1 and 2.In the case of defect X1, where a bubble is located within the lens, the resulting image shows that the white area almost merges with the black area. However, in case X2, where the bubble is indeed an air bubble, the resulting superimposed image is shown in column 5 for X2; the white area is surrounded by a black ring. The superimposition method is an improved method for detecting very fine bubbles that may be present within the lens.

[0063] Further processing takes place to detect other defects in step 108, which will not be discussed in detail. In step 110, the program checks whether any defects were found in step 108. If so, the process moves to step 109 and then to step 120 to terminate the program. If no defects were found in step 110, the program moves to step 111, where the examined contact lens is ejected before the program terminates.

[0064] The present invention is preferably used in an automated manufacturing system where the test specimen (contact lens) is transported along a predetermined path and positioned under an inspection station for inspection. Preferably, the contact lens moves through the inspection system. However, the contact lens can also be examined in a stationary position if the inspection process requires it.

[0065] In the preceding description, the invention was described with reference to specific embodiments. However, it is obvious that a person skilled in the art can make numerous modifications and changes without deviating from the broader inventive concept and scope of protection. For example, such an embodiment may consist of the use of a single-color or monochrome camera in conjunction with multiple illumination modules configured to capture multiple images of the same object at different times. Accordingly, the description and drawings are to be understood as illustrative only and not as limiting.

[0066] The method described above is suitable for the inspection of all types of contact lenses, preferably soft conventional hydrogel contact lenses comprising poly-HEMA homo- or copolymer, a PVA homo- or copolymer or a cross-linked polyethylene glycol or polysolixane hydrogel.

Claims

[1] Observation system for inspecting a contact lens (18) being transported in an automated manufacturing plant, comprising: a plurality of lighting modules (14, 15, 16, ..., N2) for illuminating the contact lens (18), which emit light, comprising various lighting techniques; a plurality of beam splitters (11, 12, 13, ..., N4) for directing the plurality of illumination to highlight the contact lens (18); a plurality of beam splitters (8, 9, 10, .... N3) for receiving a plurality of processed images of the contact lens; a plurality of filters (5, 6) for filtering the images by a suitable method; a plurality of optical modules (5, 6, 7, ..., N2) for obtaining a plurality of images of the contact lens (18), wherein each of the plurality of optical elements processes the image to enhance the contrast of different defects of the contact lens (18); a plurality of optical detectors for receiving filtered and processed images; a contact lens holder (17) for transporting the contact lens; an image analysis device for processing and analyzing the majority of images for defects in the contact lens, wherein said image analysis device is configured to extract an edge from the majority of the images; wherein said image analysis device is further configured to draw two or more concentric circles on the analyzed majority of the images and to transform an area of ​​interest encompassed by the two or more concentric circles to produce a two-dimensional image; and wherein said image analysis device is operable to analyze an inverted contact lens (18) by defining an area in the two-dimensional image by a small box, binarizing the area by defining a light intensity limit for dark and white pixels, identifying a group of white pixels in the area and evaluating the geometric property of at least one group of white pixels. [2] Observation system according to claim 1, wherein the light emitted by the illumination modules (14, 15, 16, ..., N2) can be in the i) visible spectrum, ii) infrared spectrum or the ultraviolet spectrum. [3] Observation system according to claim 1, wherein each of the beam splitters (8, 9, 10, .... N3, 14, 15, 16, ..., N2) can be a dichroic splitter. [4] Observation system according to claim 1, wherein each optical module (5, 6, 7, ..., N2) of the plurality of optical modules (5, 6, 7, ..., N2) consists of a combination of lenses. [5] Observation system according to claim 4, wherein each lens in the optical module (5, 6, 7, ..., N2) has a predetermined magnification which differs from each other. [6] Observation system according to claim 4, wherein each optical module (5, 6, 7, ..., N2) is equipped with a wavelength filter to filter suitable wavelengths of illumination. [7] Observation system according to claim 1, wherein the optical detector is a CCD monochrome camera (1, 2, 3). [8] Observation system according to claim 7, wherein each camera (1, 2, 3) is equipped with an aperture. [9] Observation system according to claim 8, wherein a larger aperture is provided for brightfield imaging. [10] Observation system according to claim 8, wherein a smaller aperture is provided to increase the depth of field. [11] Observation system according to claim 1, wherein the contact lens (18) is arranged in the contact lens holder (17). [12] Observation system according to claim 11, wherein the contact lens is suspended in saline solution or a similar liquid. [13] Observation system according to claim 2, wherein each illumination module (14, 15, 16, ..., N2) is configured to provide i) a narrow beam, ii) a diverging beam or wide-angle illumination. [14] Method for inspecting a contact lens, wherein multiple images are obtained by a plurality of optical detectors or imaging devices and subsequently processed and analyzed in the image analysis device by image processing algorithms to detect defects in the contact lens (18), the method comprising the following steps: Extracting an edge from a captured image; Drawing two or more concentric circles on the captured image and converting an area of ​​interest enclosed by the two or more concentric circles to create a two-dimensional image; and Analyzing an inverted contact lens (18), wherein the analysis steps include: a) Defining an area in the two-dimensional image using a small box b) Binarizing the area in the box by defining a light intensity threshold to separate the dark and white pixels c) Identifying the white pixel groups in the area d) Evaluating the geometric properties of at least one group of white pixels. [15] Method according to claim 14, wherein the contact lens (18) is illuminated by a plurality of lighting modules (14, 15, 16, ..., N2). [16] Method according to claim 15, wherein the light emitted by the lighting modules (14, 15, 16, ..., N2) can be in the i) visible spectrum, ii) infrared spectrum or in the ultraviolet spectrum. [17] Method according to claim 15, wherein the light emitted by the lighting modules (14, 15, 16, ..., N2) is controlled by a single trigger signal to each lighting module (14, 15, 16, ..., N2). [18] Method according to claim 17, wherein the individual trigger signal takes place for a substantially short period of time. [19] Method according to claim 14, wherein said geometric property comprises the height of at least one white pixel group. [20] Method according to claim 14, wherein said geometric property comprises the width of at least one white pixel group. [21] Method according to claim 14, wherein said geometric property comprises the image aspect ratio of at least one white pixel group. [22] Method according to claim 14, wherein said geometric feature comprises a step in which the number of white pixel groups in the horizontal orientation is counted and a further step in which the contact lens (18) is discarded when it exceeds a predetermined count result. [23] Method according to claim 14, further comprising the following steps for analyzing an inverted contact lens (18): a) Define an area around the edge in the image using a small box (Box E in Fig. 17) and (Box F in Fig. 22) in the brightfield images of an inverted or a normal contact lens (18), b) Defining an area around the edge in the image using a small box (Box C in Fig. 15) and (Box D in Fig. 20) in the dark-field images of an inverted or a normal contact lens (18), c) Overlaying the image contained in Box C onto the image contained in Box E, creating a superimposed image as in Fig. 19 is created, d) Overlaying the image contained in Box D onto the image contained in Box F, creating an overlay image as in Fig. 24 is created, e) Identifying the percentage extent of the black pixel area from the white edge of the overlay image, f) Determining the geometric property from the size ratio of black to white pixels in the overlay images that are in Fig. 19 and Fig. 24 are shown. [24] Method according to claim 23, wherein said geometric feature comprises a step in which the extent ratio of black to white pixels is compared with a programmable, user-defined value, and a subsequent step in which the contact lens (18) is discarded if the extent ratio value exceeds a predetermined ratio, as shown in the superimposed images of an inverted contact lens (18) in Fig. 19 is shown. [25] Method according to claim 23, wherein said geometric feature comprises a step in which the extent ratio of black to white pixels is compared with a programmable, user-defined value, and a subsequent step in which the contact lens (18) is passed on when the extent ratio falls into a predetermined value, as in the superimposed image of a normal contact lens (18) in Fig. 24 is shown. [26] Method according to claim 14, further comprising the following steps for analyzing a tear defect in the contact lens (18): a) Identifying the edge of the contact lens (18) in the unfolded image of a surface around the corner in Fig. 25, which is in Fig. Image 26 shows, b) Measuring the length of tears L1 and L2 in the Fig. 26 shown image, c) Evaluate the geometric properties of L1 and L2. [27] Method according to claim 26, wherein said geometric feature comprises a step in which the height of the crack L1 and L2 is measured, and a subsequent step in which the contact lens (18) is discarded if the height value exceeds the limits set by the user. [28] Method according to claim 26, wherein said geometric feature comprises a step in which the width of the tear L1 and L2 is measured and a subsequent step in which the contact lens (18) is discarded if the width value exceeds the limits set by the user. [29] Method according to claim 14, further comprising the following steps for analyzing gap defects in the contact lens (18): a) Identifying the edge of the contact lens (18) in the unfolded image in Fig. 27, which is in Fig. The image shown in image 28 shows that b) Locating a gap in the edge profile in the unfolded image of the contact lens in Fig. 28, c) Measuring the length of the gap identified in step b). [30] Method according to claim 29, wherein said geometric feature comprises a step in which the gap spacing is compared with the values ​​specified by the user, and a subsequent step in which the contact lens is discarded if the gap value exceeds the limits specified by the user. [31] Method according to claim 14, further comprising the following steps for analyzing double-edge defects in the contact lens (18): a) Identifying the edge of the contact lens (18) in the unfolded image of the image in Fig. 30, which is in Fig. The image shown in 31 shows, b) Locating a white pixel blob between two black edges in the edge profile of the unfolded image from the contact lens edge in Fig. 28, c) Measuring the geometric properties of the white pixel blob. [32] Method according to claim 31, wherein said geometric property comprises a step in which the area of ​​the pixel blot is calculated and a subsequent step in which the contact lens (18) is discarded if the pixel blot area value exceeds the limits set by the user for the double-edge defective contact lens (18). [33] Method according to claim 14, further comprising the following steps for analyzing non-circular defects in the contact lens (18): a) Identifying the edge of the contact lens (18) in the unfolded image of the image in Fig. 33, which is in Fig. The image shown in section 34 shows that b) Measure the geometric property as in vertical height at different points of the imaginary line of a normal contact lens edge line with that of the line actually identified in step a). [34] Method according to claim 33, wherein said geometric property comprises a step in which the different height values ​​measured at different points are compared and a subsequent step in which the contact lens (18) is discarded if the values ​​of the measured heights exceed the limit values ​​set by the user for a non-circularly defective contact lens (18). [35] Method according to claim 14, further comprising the following steps for analyzing the geometric property in order to distinguish between bubbles inside the contact lens (18) and air bubbles in the solution: a) Identifying white spots (X1 and X2) in a dark-field image as in Fig. 35 shown, b) Identifying the same arrangement of blots (X1 and X2) in a bright-field image as in Fig. 36 shown, c) Measuring the area of ​​the white blob in column 2 for defects X1 and X2, as shown in the table of Fig. 37 shown, d) Calculating the ratio of white pixels to dark pixels for a predetermined area around the white blob in column 2 in the Fig. Table 37 shown. [36] Method according to claim 35, wherein said geometric feature comprises a step in which the ratio of white to black pixels is compared with a user-predetermined value, and a subsequent step in which the contact lens (18) is discarded when the value of the ratio of white to black pixels does not meet a user-specified value (V2 in Fig. 38) for a blister defect inside the contact lens (18). [37] Method according to claim 35, wherein said geometric feature comprises a step in which the ratio of white to black pixels is compared with a user-predetermined value, and a subsequent step in which the contact lens (18) is accepted when the value of the ratio of white to black pixels is below a user-specified value (V1 in Fig. 38) which indicates an air bubble detected in the solution. [38] Method according to claim 35, wherein said geometric property comprises a step in which the ratio of white to black pixels is adjusted using the values ​​between V1 and V2 in Fig. 38 is compared and a subsequent step in which the contact lens (18) is categorized for further analysis if the value of the ratio of white to black pixels falls into the range between V1 and V2. [39] Method according to claim 38, further comprising the following steps for further analysis of the geometric property in order to further distinguish between bubbles inside the contact lens and air bubbles in the solution: a) Identifying white spots (X1 and X2) in a dark-field image as in Fig. 35 shown, b) Identifying the same arrangement of blots (X1 and X2) in a bright-field image as in Fig. 36 shown, c) Measuring and extracting the area of ​​the white blob in column 2 for defects X1 and X2, d) Measuring and extracting the area of ​​the black blob in column 4 for defects X1 and X2, e) Placing the white blob on top of the black blob as shown in column 5 for X1 and X2, f) Calculating the ratio of white pixels to black pixels for a given area around the white blob in column 2. [40] Method according to claim 39, wherein said geometric feature comprises a step in which the ratio of white to black pixels is compared with a programmable user-defined value, and a subsequent step in which the contact lens (18) is discarded if the value of the ratio of white to black pixels is lower than the user-defined limit for a bubble defect inside the contact lens (18). [41] Method according to claim 39, wherein said geometric feature comprises a step in which the ratio of white to black pixels is compared with a programmable user-defined value, and a subsequent step in which the contact lens (18) is accepted if the value of the ratio of white to black pixels is higher than a user-defined limit, with a decision that the defect is an air bubble in the solution and not a bubble defect inside the contact lens (18).

Images