Environmental temperature change measurement device for camera flange back

Abstract

To provide a measurement method capable of measuring a flange back fluctuation of an imaging camera with a use environmental temperature change with high accuracy of ±1 μm or less, and a device.SOLUTION: A lens 1 to be inspected is enclosed by a thermostat bath 5. An incidence window 6a is disposed in an upper part and an emission window 6b is disposed in a lower part. A first best focal position of a chart image (aerial image) at an image formation point 3 of a first chart 10a is detected from the lower part by an objective lens 12, further, a second best focal position of a chart image (direct image) on a surface of a second chart 10b is detected, and an interval distance H between the first and second best focal positions is calculated. The H with respect to each environmental temperature is measured, thereby measuring focus fluctuation ΔH and flange back fluctuation ΔFB at the image formation point 3 of the first chart defining the second chart surface as a reference. A flange face 2 of the lens 1 to be inspected and the second chart to be a position reference are connected while interposing a proximate and short interval distance L therebetween, thereby being extremely hard to be influenced by disturbance.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to a measuring instrument and an inspection instrument for assuring flange focal distance deviations due to environmental temperatures of photographic cameras such as in-vehicle cameras, mobile phone cameras, and CCTV cameras. [Background technology]

[0002] As an example of a photographic camera, the structure of a typical vehicle-mounted camera is shown in Figure 1. It consists of a photographic lens unit 100 that captures light beams from a target object and forms an image, and a camera unit 14 that converts the image photographed on an image sensor 15 into an electrical signal on a circuit board 16, performs image processing (exposure correction, aberration correction, color correction, etc.) as necessary, and outputs the signal to an external device.

[0003] If the ambient temperature fluctuates, the refractive index, radius of curvature, lens thickness, and lens spacing of each lens in the lens unit will change, which may result in focus shifts, degradation of resolution, and deterioration of other optical aberrations (distortion, field curvature, etc.).

[0004] The temperature characteristics of the entire camera unit can be evaluated by placing the camera in a thermostatic chamber and inspecting the quality (resolution, image distortion, etc.) of the output image captured using a resolution chart or similar. However, since the optical manufacturer that produces the lens unit 1 and the electronics manufacturer that produces the camera unit 3 are typically in different technical fields, shipping and acceptance inspections that guarantee the quality of the lens unit alone are necessary. Normally, imaging characteristics are guaranteed at room temperature. However, focus deviation due to environmental temperature has a significant impact on imaging characteristics, and cameras used in environments with large fluctuations in outside temperature, such as in-vehicle cameras, mobile smartphone cameras, and V (surveillance cameras), are often designed to minimize focus deviation, requiring temperature assurance of the lens unit alone.

[0005] From a security standpoint, in-vehicle cameras in particular require safety assurance based on extremely strict standards such as the international standard ISO26262, the automotive industry SQMS standard TS16949, and cybersecurity (CS) regulations, and there is a demand for guaranteed temperature characteristics for the lens unit alone in environments ranging from extremely low temperatures of -40°C to high temperatures of +125°C.

[0006] The so-called flange back determines the focal position of the photographic lens unit and the camera sensor. As shown in Figure 1, the flange back is the distance FB from the flange surface 2 of the lens unit barrel to the surface of the image sensor 15 (taking into account the amount of optical lift of the cover glass, IR filter, etc.).

[0007] When environmental changes occur, it is important to suppress fluctuations in flange focal length. Generally, fluctuations in flange focal length (imaging focus) are suppressed using the following methods. 1) A method of optical design in which fluctuations in the temperature coefficients (refractive index, linear expansion coefficient, etc.) of each lens that makes up a lens unit, and in the power (focal length) of each lens and each lens group cancel each other out has been proposed and adopted in JP 2016-170446, JP 2017-198947, JP 2021-156953, etc. 2) In addition to the above-mentioned measure 1), methods of optimizing the structure and linear expansion coefficient of the lens frame, support members, etc. to cancel fluctuations in the focus position of the entire camera unit have been proposed and adopted in JP 2016-218139 A, JP 2019-197231 A, JP 2022-120334 A, etc.

[0008] Using the methods described above, flange back shift and focus shift due to changes in environmental temperature are corrected, but until now it has been difficult to evaluate flange back shift and temperature characteristics (MTF, etc.) of a single lens unit through actual measurements, and it has been necessary to guarantee design values ​​or optical simulation guarantees of temperature characteristics.This method only guarantees based on optical simulations and simulations of lens barrel expansion and contraction, so it has not been possible to verify through testing whether focus shift has actually been suppressed as predicted by the simulations.

[0009] In particular, from the perspective of national security, automotive cameras are required to meet extremely strict standards, such as the international standard ISO26262, the automotive industry QMS standard TS16949, and cybersecurity (CS) regulations, and are required to guarantee focus deviation of less than a few microns over a temperature range from the frigid cold of -40°C to the maximum temperature inside a vehicle of approximately 125°C. Furthermore, in recent years, the pixel size of image sensors has become increasingly fine, requiring higher resolution from imaging lens units. This has led to a trend toward larger numerical apertures (NA), which in turn has led to a shallower (smaller) depth of field, which is the tolerance for focus deviation. Therefore, it has become necessary to limit flange back deviation of a single lens unit due to environmental temperature fluctuations to approximately ±3 μm.

[0010] It is extremely difficult to guarantee compliance with this standard through simulation alone, and there is a demand for a method and device that can accurately measure environmental fluctuations within a temperature range of approximately -40°C to +125°C for an in-vehicle camera and flange focal length fluctuations within ±1μm. [Prior art documents] [Patent documents]

[0011] [Patent Document 1] Patent Publication No. 2016-170446 [Patent Document 2] Japanese Patent Application Publication No. 2017-198947 [Patent Document 3] Japanese Patent Publication No. 2021-156953 [Patent Document 4] Japanese Patent Application Laid-Open No. 2016-218139 [Patent Document 5] Japanese Patent Application Publication No. 2019-197231 [Patent Document 6] Japanese Patent Publication No. 2022-120334 [Patent Document 7] Japanese Patent Application Publication No. 10-68674 [Patent Document 8] Japanese Patent Application Laid-Open No. 2003-50183 [Patent Document 9] Japanese Patent Application Laid-Open No. 2004-294076 [Patent Document 10] Japanese Patent Application Publication No. 2019-49524 [Patent Document 11] Japanese Patent Publication No. 2022-191392 Summary of the Invention [Problem to be solved by the invention]

[0012] The resolution performance of a camera is generally tested by measuring the modulation transfer function (MTF). Many methods and devices for testing MTF have been proposed and are on the market, and can be broadly divided into orthogonal projection and inverse projection methods.

[0013] The orthogonal projection method places a chart on the object side of the lens under test and outputs the MTF value from the chart image formed by the lens under test. This method has been proposed in Japanese Patent Application Laid-Open Nos. 10-68674, 2003-50183, and 2004-294076, among others. As will be described later, the orthogonal projection method has the advantage of being able to inspect high-frequency resolution and making high-efficiency use of the light source (halogen, etc.) of the measuring instrument. On the other hand, when measuring MTF over a wide angle of view, it is necessary to move the image sensor or use a large sensor. Therefore, when prioritizing the number of inspections and shortening the inspection time, the back projection method is considered to be slightly more advantageous.

[0014] The back projection method, which involves placing a chart on the image side of the lens under test and outputting the MTF value from the chart image formed on the object side by the lens under test, has been proposed in JP 2019-49524 A and JP 2022-191392 A, among others. Compared to the orthogonal projection method, the back projection method forms a chart corresponding to each angle of view on the image plane, allowing for nearly simultaneous measurement of MTF values ​​for multiple angles of view, which has advantages in terms of the number of items to be inspected and inspection time. However, because the line width of the chart determines the measurement resolution, it is difficult to measure MTF at frequencies (100 lines / mm or more) when using a line width of 5 μm, which is currently guaranteed.

[0015] Although several devices capable of measuring the temperature characteristics of MTF have been proposed, there is almost no appropriate documentation available in patent documents, papers, etc. Therefore, an example using the orthogonal projection method is shown in Figure 2. The lens 1 to be tested is surrounded by a thermostatic chamber, the outer walls of which are covered with insulating material to maintain as much thermal insulation as possible from the outside air, and the temperature inside the thermostatic chamber is varied by introducing temperature-controlled gas or air from low to high temperatures.

[0016] The resolution chart 10 is illuminated by a light beam from the light guide 7, converted into a nearly parallel light beam by the collimator lens 11, and then focused by the test lens 1 to form a chart image at the imaging point 3, constituting a so-called orthogonal projection optical system. This chart image is then re-formed on the image sensor 15 via the microscope objective lens 12. The MTF value can be calculated from the contrast of this image or by using a Fourier transform, etc. Furthermore, by measuring the MTF while defocusing the microscope objective and image sensor in the optical axis direction, the so-called MTF-with-focus characteristics (defocus vs. MTF characteristics) can be measured, and the position with the highest MTF can be measured as the just-in-focus position.

[0017] If the same focus position is measured at each temperature, it is theoretically possible to measure the fluctuation in the focus position, i.e., the fluctuation in the flange back shift. However, the flange surface 2 of the lens 1 under test, the objective lens 12, and the image sensor 15 are not directly connected, but are connected via multiple members in between, namely the lens holder 4, the thermostatic chamber 5, the device base 21, and the automatic stage 20, with a substantial distance L between them.

[0018] Each component is made of materials with various coefficients of linear expansion, and even a change of a few degrees in the ambient room temperature during measurement can cause significant shifts in the imaging position. This makes it impossible to accurately measure temperature fluctuations in the flange focal distance. Furthermore, the lens holder 4 in the thermostatic chamber changes directly from low to high temperatures, causing large shifts in the imaging point due to thermal expansion and contraction. Furthermore, because the tip of the objective lens is located below the thermostatic chamber, the focal length of the microscope becomes unstable due to the influence of the thermostatic chamber temperature. Furthermore, because the incident light side (the incident light beam from the collimator) and the exit side of the thermostatic chamber are not completely isolated from the outside temperature, air fluctuations due to temperature differences occur, significantly degrading the MTF measurement accuracy and the accuracy of measuring the flange focal distance's exact focus position.

[0019] As described above, it was difficult to accurately measure the temperature fluctuations of the MTF and flange focal length using the conventional MTF temperature characteristic measuring device shown in FIG.

[0020] Therefore, a back-projection type MTF temperature characteristic evaluation device has been proposed and sold. This method is also not clearly disclosed in documents such as patent documents, so an overview of the method is shown in Figure 3.

[0021] The lens 1 to be tested is surrounded by a thermostatic chamber 5, the outer walls of which are covered with insulating material to maintain as much thermal insulation as possible from the outside air. A glass entrance window is installed on the incident light beam side, and temperature-controlled gas or air can be injected to maintain a constant temperature inside the thermostatic chamber, ranging from low to high. The resolution chart 10 is a so-called back-projection optical system installed near the image position of the lens 1 to be tested. Here, the chart 10 is usually a component with a cross slit or similar formed on a glass plate, so even if large temperature fluctuations occur inside the thermostatic chamber, from extremely low to high temperatures, it does not affect the deterioration of optical aberrations. This makes it possible to place the chart inside the thermostatic chamber.

[0022] The chart 10 is irradiated with a light beam from the light guide 7, passes through the lens under test, and forms a chart image on the image sensor 15 by the collimator lens 11. The temperature characteristics of the MTF value of this chart image can be evaluated by the contrast method or Fourier transform method.

[0023] Using this MTF device, it is possible to evaluate the temperature characteristics of the flange focal length. By measuring the MTF while defocusing the chart 10 in the optical axis direction, the so-called MTF under-focus characteristics (defocus vs. MTF characteristics) can be measured, and the position with the highest MTF value can be measured as the just-in-focus position. By measuring the same focus position at each temperature, it is in principle possible to measure the fluctuation in the focus position, i.e., the fluctuation in the flange focal length deviation.

[0024] As described above, this method is currently considered to be the best method for evaluating the best MTF and temperature characteristics of flange back.However, when measuring flange back shift (focus shift), the following error factors are large, and in reality, measurement accuracy can reach approximately ±10 μm, which poses a major issue: it is not possible to obtain the measurement accuracy required to guarantee flange back shift at the level of ±3 μm that is currently required.

[0025] The following are some of the factors that cause errors when measuring flange back deviations due to environmental temperatures using a conventional MTF measuring device. a. Temperature fluctuation error from the test lens FB reference surface to the chart surface: Approximately ±5 μm b. Detection error of just-in-focus position due to MTF through-focus characteristics: Approximately ±2 μm c. Temperature distribution stability in the thermostatic chamber: Approximately ±1 μm d. Automatic stage feed accuracy: Approximately ±1 μm Here, c. can be improved to approximately ±0.2 μm by improving the thermal insulation structure inside the thermostatic chamber, etc. (The details of this improvement method are not described in this patent specification.) The above d. can be improved to approximately ±0.1 μm by using a high-resolution encoder. Here, we will explain the important error factors a. and b.

[0026] 3 of the conventional example, the distance from the flange surface 2, where the test lens 1 is abutted, to the image formation point 3 is the so-called flange back distance FB. When measuring the flange back shift in response to changes in environmental temperature, the chart 10 is defocused in the optical axis direction with the automatic stage 20 when the temperature is changed, the MTF through-focus characteristics are measured, and the position with the highest MTF value (highest resolution) is taken as the just-in-focus position.

[0027] Next, while changing the temperature from low to high, when a constant temperature is reached at any desired temperature, the just-in-focus position is determined in the same way, and the change in focus at each temperature is measured as flange back shift. However, the flange surface 2 of the lens 1 under test and the chart 10 are not directly connected, but are connected via a number of components, including the lens holder 4, the thermostatic bath 5, the device base 21, the automatic stage 20, and the illumination optical system 30, with a substantial distance L between them.

[0028] Each component is made of materials with various coefficients of linear expansion, and even a change of a few degrees in the room temperature during measurement can cause a significant shift in the imaging position. Furthermore, the lens holder 4 in the thermostatic chamber and the chart holder installed in the thermostatic chamber change directly from extremely low to high temperatures (-40°C to 125°C), so thermal expansion and contraction cause a large shift in the imaging point. Of the factors that cause errors in measuring flange back deviation, a. described in

[0025] occurs at a rate of approximately ±5 μm or more.

[0029] Next, we will explain the cause of errors in detecting the in-focus position from the MTF through-focus characteristics (b) described in

[0025] , which is a major cause of measurement errors, with reference to Figure 4. When measuring MTF while defocusing and detecting the true in-focus position A from the MTF through-focus characteristics, electrical noise such as camera spike noise and dark current noise, as well as mechanical and optical noise such as vibrations during automatic stage movement and air fluctuations, are generated. The influence of these noises makes it difficult to accurately determine the in-focus position, particularly in optical systems with a shallow depth of focus, and the detected in-focus position B may be off from the true in-focus position. Therefore, averaging processing is usually performed to reduce the effects of noise to some extent, but conventional methods still produce errors of about ±2 μm.

[0030] Furthermore, because the conventional method shown in Figure 3 uses a back projection method, the maximum frequency depends on the line width of the chart. Ordinary charts are produced in the same way as semiconductor manufacturing processes, by depositing a film of chromium or other material on a glass substrate, then exposing and etching it to form a cross chart or edge chart, but the guaranteed accuracy of line width and edge is generally around 5 μm. Therefore, the maximum frequency Fm for MTF measurement is limited to around 100 lines / mm (Fm = 1 / [line width × 2] = 1 / (0.005 mm × 2)).

[0031] Due to the recent demand for smaller and higher density imaging elements, pixel sizes of around 2 μm are becoming commonplace, and the evaluation frequency for image resolution, i.e., the Nyquist frequency Fn, is required to be evaluated at 250 lines / mm (Fn = 1 / [line width × 2] = 1 / (0.002 mm × 2)), and the inability to evaluate this has become a problem.

[0032] On the other hand, with the orthogonal projection method, as shown in the conventional example in Figure 2, the chart image passes through collimator 11 and is formed on test lens 1. However, because the optical magnification of the collimator and test lens is typically 10x or more, the chart image on the image plane can be formed at a frequency 10x higher, and theoretically, frequencies up to 1000 lines / mm can be measured. In practice, the orthogonal projection method makes it possible to measure MTF values ​​up to approximately 300 lines / mm due to constraints such as the number of bits and noise in the imaging signal. However, as mentioned above, evaluation of MTF temperature characteristics using the conventional orthogonal projection method has the problem of low MTF measurement accuracy because the microscope objective is affected by the temperature of the thermostatic bath.

[0033] From the above, the challenge with conventional MTF temperature characteristic measuring devices is that in order to guarantee the new specification of flange back deviation against environmental temperature of ±3μm or less, it is necessary to establish a flange back measurement accuracy of ±1μm or less. Furthermore, there is a demand for the realization of an MTF and flange back measurement method and device that can measure MTF frequency characteristics up to about 300 lines / mm. [Means for solving the problem]

[0034] An object of the present invention is to provide a method and apparatus that can measure defocus with high accuracy and ensure high-frequency MTF even in a wide temperature range from low to high.

[0035] Figure 5 shows the overall configuration of the present invention. As shown in Figure 5, the basic structure of the present invention is an orthogonal projection type MTF measuring instrument, and its greatest feature is the use of two or more charts: a first chart 10a on the illumination light source side and a second chart 10b near the image point of the lens under test. The light beam from the first chart is collimated by a collimator 11, then enters the lens under test 21 and forms an image. The distance FB from the flange surface, which is the abutting surface of the lens under test, to the image point 3 is the so-called flange back. Here, when evaluating MTF performance at a finite distance, a finite lens (a lens whose object point is not at infinity but at a finite distance) may be used instead of the collimator 11.

[0036] The test lens 1 is surrounded by a thermostatic chamber 5, and has an entrance window 6a at the top and an exit window 6b at the bottom. From the bottom, an objective lens 12 photographs a chart image of the image-forming point 3 of the first chart 10a, and also photographs a chart image of the surface of the second chart 10b directly.

[0037] "Explanation of Claim 1" 5, the objective lens 12 is moved in the optical axis direction to detect the best imaging position 3 (first position) of the first chart 10a and the best focus position (second position) on the surface of the second chart 10b, and the distance H between the first position and the second position is calculated. Here, as the first position detection means for detecting the best imaging position 3 of the first chart 10a and the second position detection means for detecting the best focus position on the surface of the second chart 10b, the first and second images can be directly visually confirmed, but for highly accurate detection, it is effective to measure the MTF through-focus characteristics as shown in FIG.

[0038] The MTF of each image can be calculated by using the contrast method or Fourier transform on the chart image (aerial image) of the first chart 10a and the surface image (direct image) of the second chart 10b. By calculating the MTF while moving the objective lens 12 in the optical axis direction, the MTF of the first chart image and the second chart image can be observed, as shown in Figure 8, and the so-called MTF through-focus characteristics can be measured. The MTF peak values ​​of the first and second charts, i.e., the in-focus positions of each chart image, are measured, and the distance H between the peaks is calculated. By measuring H for each ambient temperature, the focus variation ΔH and flange back variation ΔFB of the image point 3 of the first chart relative to the surface of the second chart can be measured. Fig. 9 shows a measurement example of flange back variation ΔFB as a function of the ambient temperature.

[0039] Here, we explain the reason for using the second chart. In the conventional examples shown in FIGS. 2 and 3, the flange surface of the lens 1 under test and the detection reference (the objective lens in FIG. 2, and the chart in FIG. 3) are not directly connected, but rather are connected via numerous components, with a substantial long distance L between them. This results in flange focal length measurement errors of ±5 μm or more, as described in

[0025] . In contrast, in the present invention, shown in FIG. 5, the flange surface 2 of the lens 1 under test and the second chart serving as the position reference are connected closely to each other with a short distance L between them, making them extremely resistant to the effects of external disturbances. Therefore, even under a wide temperature range from extremely low (-40°C) to high (125°C) in a thermostatic chamber, the invention described below can significantly reduce flange focal length measurement errors to around ±0.1 μm.

[0040] In the above example, two resolution charts, the first chart and the second chart, are used. However, it is also possible to use multiple charts for each of multiple angles of view in the first chart section, and measure the focus variation ΔH of the imaging point and the flange back variation ΔFB for each angle of view.

[0041] "Explanation of Claim 2" In the conventional method, a microscope objective lens is used as the objective lens 12, as shown in Figure 2, and the normal working distance is approximately 5 mm or less. Therefore, the tip of the objective lens is configured to be inserted into the bottom of the thermostatic chamber. As a result, the optical performance of the objective lens (focal length, etc.) is significantly affected by changes in the temperature of the thermostatic chamber, and air fluctuations are likely to occur near the objective lens, causing distortion in the formed chart image. This reduces the accuracy of the MTF measurement and makes it impossible to accurately detect the best imaging position. 5, the present invention is characterized in that an exit window 6b is provided in the thermostatic chamber 5 to thermally insulate the inside of the thermostatic chamber from the outside air, and the best imaging position 3 (first position) of the first chart and the best focus position (second position) of the surface of the second chart are observed from the outside through this exit window with the objective lens 12, and by using a microscope objective lens with a long working distance (WD) for this objective lens, the best imaging position 3 (first position) of the first chart and the best focus position (second position) of the surface of the second chart are observed. This configuration improves the temperature stability inside the thermostatic chamber, and the objective lens is not affected by the temperature of the thermostatic chamber, allowing the first and second positions to be detected with high accuracy while maintaining stable optical characteristics (focal length, etc.).

[0042] As described above, by installing an exit window such as a glass plate at the bottom of the thermostatic chamber, the thermostatic chamber's heat-insulating effect is improved, improving the temperature stability inside the chamber and thermally isolating it from the objective lens. In the present invention, as shown in FIG. 5, an entrance window 6a is also provided on the entrance side. Furthermore, by using a double-glass structure for the entrance and exit windows, condensation prevention and heat insulation effects can be further enhanced. This also means that air fluctuations caused by temperature fluctuations are almost completely eliminated. This also improves the accuracy of MTF measurement and the accuracy of detecting the flange back focal position.

[0043] The microscope objective lens with the long working distance described above is designed to correct the spherical aberration caused by the glass thickness of the exit window. If the substrate of the second chart is thick and the captured image is degraded, an objective lens that is optically designed to also correct the spherical aberration caused by the substrate thickness is used. The temperature tends to be slightly unstable up to approximately 5 mm inside the thermostatic chamber near the entrance and exit windows, and since it is necessary to move the objective lens along the optical axis by several millimeters or more to find the best imaging position, it is desirable to optically design the objective lens so that the working distance WD is 10 mm or more.

[0044] "Explanation of Claim 3" Figure 6 of the present invention is an enlarged, more detailed illustration of the vicinity of the lens under test in Figure 5. In Figure 6, a second chart 10b is provided in addition to the first chart 10a, which is the greatest feature of the present invention, to explain the method of measuring flange back deviation in response to changes in environmental temperature. The first chart 10a is converted into nearly parallel light by a collimator 11, then focused by the test lens 1 to form a chart image at the focusing point 3. Here, the distance from the flange surface 2 to the image focusing point 3 is the flange back FB. On the other hand, the second chart is located near the focusing point 3. As shown in Figure 7(a), the first and second charts used here are as follows: the first chart 10a is a transmission chart with a negative cross slit formed on a glass substrate, and is shielded by a light-shielding film such as a chrome film with an optical density (OD) of 3 or more so that no light beam passes through except the cross slit; the second chart 10b is a transmission chart with a positive cross slit formed on a glass substrate with an optical density (OD) of 3 or more.

[0045] When the two chart images are captured with the microscope objective lens 12, a bright cross image (aerial image) of the first chart and a dark cross image (direct image of the chart surface) of the second chart are observed, as shown in Figure 7(b). Because the second chart is transmitted through the objective lens except for the cross slit, the objective lens can simultaneously observe the cross image formed on the first chart and the cross image on the surface of the second chart through the second chart. The MTF of each cross image can be calculated by extracting each image of the two cross images and performing computations such as the contrast method or Fourier transform processing. While the method described here uses simultaneous observation of the two chart images, if image extraction is difficult due to the cross images of the first chart and the second chart being close to each other on the image sensor, it is also possible to calculate the MTF of the first chart and the second chart separately by switching on either the light source of the light guide 7 for the first chart or the light source of the illumination optical system 31 for the second chart (see Figure 5).

[0046] All of the above charts are manufactured using the same methods as semiconductor manufacturing processes (film formation, exposure, etching, etc.), but the minimum slit line width currently available is about 5 μm. While the present invention is based on measuring MTF and flange focal distance using an orthogonal projection optical system, the focal length of the collimator 11 is selected to provide a magnification of 10 or more times that of the test lens 1 to improve the measurement accuracy of high-frequency MTF. Therefore, the maximum measurable MTF frequency of the first chart is 10 times or more the chart line width, enabling evaluation at a maximum MTF measurement frequency of 300 lines / mm or more, as described in

[0032] . With an optical magnification of 10x, MTF evaluation is possible at a maximum of 1000 lines / mm using a 5 μm line width, and 100 lines / mm even using a 50 μm line width chart. Incidentally, with a back projection optical system, as described in

[0030] , the maximum frequency Fm for MTF measurement is limited to approximately 100 lines / mm. Therefore, it is desirable for the resolution chart used in the present invention to have a line width of approximately 5 μm to 50 μm.

[0047] "Explanation of Claim 4" In the conventional example shown in FIGS. 2 and 3, as described in

[0039] , the flange surface of the lens under test 1 and the detection reference (the objective lens in FIG. 2, and the chart in FIG. 3) are not directly connected, but are connected via numerous components, with a substantial long distance L separating them. This results in flange back measurement errors of ±5 μm or more. In contrast, in the present invention, as shown in FIG. 6, the flange surface 2 of the lens under test 1 and the second chart 10b serving as the positioning reference are connected closely with a short distance L separating them. Furthermore, the lens holder 4, which is installed in contact with the flange surface 2 of the lens under test 1, is characterized in that the member on which the second chart 10b is installed is made of the same material and directly connected. Therefore, even under a wide temperature range from extremely low (-40°C) to high (+125°C) in the thermostatic chamber, the variation in the distance L between the flange surface 2 and the second chart surface is small, and the variation is smooth due to the use of the same components and materials.

[0048] "Explanation of Claim 5" Although the distance between the flange surface and the second chart of the lens holder 4 according to the present invention shown in Fig. 6 is shorter than that of the conventional method, expansion and contraction due to changes in environmental temperature will result in measurement errors of the flange back, so by using a material with a small linear expansion coefficient, it is possible to reduce errors caused by expansion and contraction. For example, if the distance between the flange surface and the second chart is set to L = 10 mm, and an ordinary iron material has a linear expansion coefficient α = 12 × 10 -6 When a temperature change of ΔT=100°C is made from room temperature (25°C) to a high temperature of 125°C using a lens holder 4, the lens holder 4 elongates by ΔL=12 μm (=L×α×ΔT). This elongation causes a large measurement error when measuring the temperature change ΔFB of the flange focal length.

[0049] Therefore, in the present invention, the linear expansion coefficient α is 3.0 × 10 -6 / K or less. Materials that correspond to this linear expansion coefficient include quartz: 3.0 × 10 -6 / K, Silicon: 2.6 x 10 -6 / K, Super Inverter: 1.3 x 10 -6Although the materials are limited, there are several candidate materials, such as / K. For example, if the distance between the flange surface and the second chart is set to L = 10 mm, and a normal super inverter is used, and the temperature is changed from room temperature (25°C) to a high temperature of 125°C by ΔT = 100°C, the change in elongation of the lens holder 4 can be kept to ΔL = 1.3 μm (= L × α × ΔT), which is an error that can be kept small enough to measure and evaluate the temperature characteristics of a normal flange back.

[0050] "Explanation of Claim 6" By implementing the above-mentioned claims 5 and 6 of the present invention, it is possible to significantly reduce measurement errors in flange back fluctuations that accompany expansion and contraction due to temperature fluctuations in the components that make up the measuring instrument compared to conventional methods. However, in recent years, flange back fluctuations of approximately ±3 μm in response to environmental temperature fluctuations (-40°C to +125°C) have come to be required, and it is therefore necessary to achieve a measurement accuracy of at least ±1 μm or less for the temperature characteristics of flange back.

[0051] Therefore, the present invention is characterized in that the true flange back temperature variation ΔFBr is obtained by correcting the expansion / contraction component ΔL due to temperature variation of the distance L from the flange surface of the lens to the surface of the second chart with respect to the flange back temperature variation ΔFB obtained in claim 5 or claim 6. When the distance from the flange surface to the surface of the second chart is L and the linear expansion coefficient of the lens holder is α, the expansion / contraction correction amount ΔL of the lens holder is calculated as follows: (Math. 1) ΔL=L×α×ΔT This ΔL becomes an error component when measuring the flange back variation ΔFB. Next, when the distance between the image point of the first chart and the surface of the second chart is H and the temperature fluctuation of that distance is ΔH, the true flange back temperature fluctuation ΔFBr can be calculated as follows: (Equation 2) ΔFBr=ΔH-ΔL

[0052] The above calculation procedure will be explained in a little more detail. If the distance obtained from the first position detection means and second position detection means at a reference temperature (usually room temperature) is Hs, and the distance from the flange surface of the lens under test to the position of the second chart surface is L, the flange back FBr at the reference temperature is given by the following formula. Here, L can be measured as an absolute value of the flange back if it is measured in advance using a step gauge (ideally with an accuracy of ±1 μm or less). However, when measuring and evaluating only the temperature fluctuations of the flange back, the measurement accuracy of L is not important. (Math 3) FBr=L-Hs Here, if L is measured in advance using a step gauge or the like, the absolute value of the flange back can also be measured. Next, when the distance from the flange surface to the second chart surface is L and the linear expansion coefficient of the lens holder is α, the expansion / contraction correction amount ΔL of the lens holder is calculated as follows: (Math. 1) ΔL=L×α×ΔT Next, the flange back FBt when a temperature change of ΔT occurs from the reference temperature can be calculated using the following formula. (Math. 4) FBt=L-ΔL-Ht The true flange back variation ΔFBr with respect to changes in environmental temperature can be calculated using the following formula. (Math. 2) ΔFBr=FBt-FBr=(L-ΔL-Ht)-(L-Hs) =(Hs-Ht)-ΔL=ΔH-ΔL This new flange back measurement method is named the "twin chart method."

[0053] "Explanation of Claim 7" When measuring the temperature variation of the flange focal length FB using the conventional method, as described in

[0025] , when detecting the just-in-focus position from the MTF through-focus characteristics, an error of about ±2 μm occurs due to the influence of electrical noise such as camera spike noise and dark current noise, as well as mechanical and optical noise such as vibration during movement and air fluctuation. Here, in order to detect the just-in-focus position from the MTF through-focus characteristics with high accuracy, the present invention is characterized by detecting the MTF peak position, i.e., the just-in-focus position of the chart image, according to the flowchart shown in FIG.

[0054] Figure 11 shows an image of the MTF through-focus characteristic data after processing it according to the follow chart in Figure 10. First, when measuring the MTF through-focus characteristic, the measurement frequency is determined. Generally, measuring MTF at a higher frequency will result in a sharper, more pointed MTF through-focus, as shown in Figure 11(b), improving the accuracy of detecting the MTF peak position. The absolute value of the flange back distance may vary slightly depending on the frequency, but the temperature fluctuation ΔFB is hardly affected by the frequency.

[0055] Next, when calculating the MTF at each defocus position during MTF through-focus measurement, the effects of camera noise can be reduced by capturing images several times and averaging them. Furthermore, noise components can be reduced by performing noise filtering using a low-pass filter, band-pass filter, etc. Furthermore, noise can be further reduced by fitting an approximate curve to the MTF through-focus characteristic curve. By performing the above noise processing, it is possible to significantly remove noise components, as shown in Figure 11(c).

[0056] The position where the MTF of the MTF through-focus data after the noise reduction process is maximized may be determined as the in-focus position. However, if small undulations remain in the MTF through-focus characteristic, sufficient repeatability may not be achieved. In such cases, reproducibility may be improved by setting a threshold value and determining its center point as the in-focus position. Here, as shown in Figure 11(d), for example, by setting the threshold value to 80% of the maximum MTF value and determining the center point between the left and right intercepts of the MTF through-focus characteristic as the in-focus position, it is possible to determine the in-focus position with high repeatability even when there are undulations near the top of the through-focus characteristic curve. The threshold value should be set based on the shape of the MTF through-focus characteristic curve. It is desirable for both ends of the curve to begin to change sharply; therefore, it may be set at a value other than 80% as in the above example.

[0057] It is not necessary to perform all of the above data processing steps, but by combining these data processing steps, it is possible to reduce the error in the detection accuracy of the just-in-focus position, which in the conventional method had an error of about ±2 μm, to about ±0.5 μm with the method of the present invention. [Effects of the Invention]

[0058] The measurement error of the temperature fluctuation of the flange back when the present invention is implemented as described above will be compared with the measurement error of the conventional example described in

[0025] . The error factors when measuring the flange back deviation with a normal MTF measuring device are as follows: a. Fluctuation error of the chart surface from the test lens FB reference surface By adopting the new "twin chart method", the accuracy can be significantly improved from the previous accuracy of approximately ±5μm to approximately ±0.1μm. b. Error in detecting the just-in-focus position from the MTF through-focus characteristics The previous accuracy of approximately ±2 μm can be significantly improved to ±0.5 μm by using the data processing described above. c. Stability of temperature distribution in the thermostatic chamber By improving the insulation structure inside the thermostatic chamber, the temperature distribution inside the chamber is controlled to within approximately 2°C. It has been confirmed that the accuracy can be improved from the previous level of approximately ±1 μm to approximately ±0.2 μm (the details of this improvement method are not described in this patent). d. Stage feed accuracy: By using a high-resolution encoder, the feed accuracy can be improved from the previous level of approximately ±1 μm to ±0.1 μm. From the above, it is possible to measure the overall flange back temperature fluctuation with an error of ±1 μm or less.

[0059] Due to the demand for higher pixel size and higher resolution in recent image sensors, the depth of focus, which is the tolerance for flange focus deviation, has tended to become shallower, making it necessary to suppress flange back deviation to approximately ±3 μm. It is extremely difficult to guarantee this accuracy through simulation alone, and there is a demand for a method and device that can accurately measure flange back deviation to within ±1 μm or less across a temperature range of approximately -40°C to +125°C for an in-vehicle camera, but this invention can significantly reduce the measurement error in flange back deviation to within ±1 μm, making it possible to guarantee the quality of the lens unit itself.

[0060] Furthermore, the conventional method shown in Figure 3 uses a back projection method, so it was limited to evaluating MTF at frequencies up to about 100 lines / mm. The present invention uses an orthogonal projection method to achieve temperature characteristics, and as described in

[0046] , it is possible to evaluate MTF measurements at a maximum frequency of 300 lines / mm or more. Therefore, the present invention can provide a method and apparatus that can measure high-frequency MTF and ensure high-precision measurement of focus deviation even in a wide temperature range, from low to high temperatures. [Brief explanation of the drawings]

[0061] [Figure 1] FIG. 1 is a structural diagram of a typical vehicle-mounted camera. [Figure 2] This is a conventional example of an MTF temperature characteristic measuring device that uses the orthogonal projection method. [Figure 3] This is a conventional example of an MTF temperature characteristic measuring device that uses the back projection method. [Figure 4]10A and 10B are diagrams illustrating factors that cause errors in detecting a just-in-focus position according to a conventional method. [Figure 5] FIG. 1 is a diagram illustrating the configuration of an MTF and flange focal length measuring device according to the present invention. [Figure 6] This is an enlarged view of the vicinity of the lens to be tested in FIG. 5, showing it in more detail. [Figure 7] 1 illustrates the shapes of two types of resolution charts used in the present invention and the chart images on the image sensor. [Figure 8] FIG. 10 is an explanatory diagram of a method for determining a flange back variation from the interval between the best positions of two or more chart images used in the present invention. [Figure 9] FIG. 10 is a diagram showing a measurement example of a flange back variation ΔFB with respect to the ambient temperature. [Figure 10] FIG. 1 is a flowchart for detecting the just-in-focus position with high accuracy from the MTF through-focus characteristic. [Figure 11] This shows the MTF through-focus characteristics when the processing of FIG. 10 is performed. [Explanation of symbols]

[0062] 1. Test lens 2 Flange surface 3. Image point 4 Lens holder, 5 constant temperature bath, 5a constant temperature gas inlet, 5b constant temperature gas outlet, 6 thermostatic chamber window, 6a entrance window, 6b exit window 7 Light Guide 8 Filters 9 Diffuser 10 Chart, 10a 1st Chart, 10b 2nd Chart, 11 Collimator 12 Objective Lens 13 Imaging lens 14 Camera Department 15 Image Sensor 16 Circuit Board 17 Connectors 20 Automatic Stage 21 Device base 25 Test lens adjustment mechanism; 30 Illumination optical system 31 Second chart illumination optical system 100 Photographic lens unit DETAILED DESCRIPTION OF THE INVENTION [Example]

[0063] The following shows examples of the inventions of claims 4, 5, and 6. In Fig. 6, the lens holder, which is placed in contact with the flange surface of the lens under test, and the member on the surface of the second chart are directly connected and made of the same material (claim 4). Here, the material of the lens holder under test is made of a material with a low linear expansion coefficient, such as superinvar (thermal expansion coefficient 1.3 x 10 -6 ) is used (Claim 5). For example, if the distance L between the flange surface and the second chart is set to 10 mm and the temperature is changed from room temperature (25°C) to a high temperature of 125°C by ΔT=100°C, the lens base 4 will expand and contract by ΔL=1.3 μm (=L×α×ΔT). Since the lens base 4 is made of the same material and exhibits highly isotropic expansion and contraction, if the temperature distribution in the thermostatic chamber is stable, the lens base 4 will expand and contract proportionally according to the theoretical value. Therefore, by correcting (Equation 1) and (Equation 2), even if the expansion and contraction deviates from the theoretical value by 10%, the temperature characteristics of the flange back can be measured and evaluated with an accuracy of about ±0.1 μm. This represents a significant improvement in accuracy compared to the conventional measurement error of

[0025] a. Temperature fluctuation error from the test lens back surface to the chart surface: approximately ±5 μm.

Claims

1. A method or measurement device for measuring the modulation transfer function (MTF value) of a test lens versus environmental temperature, using two or more resolution charts, an orthogonal projection optical system in which at least one of the resolution charts is placed on the object side of the test lens, and having a first position detection means for detecting the best imaging position, and another resolution chart is placed near the imaging position of the test lens and has a second position detection means for detecting the chart surface position, and measuring the difference distance obtained from the first position detection means and the second position detection means for each environmental temperature to measure flange back variation.

2. 2. The method and device for measuring MTF and flange back variation according to claim 1, wherein a thermostatic chamber is used as the means for varying the environmental temperature, a test lens is placed inside the thermostatic chamber, the test lens has an entrance window for making a light beam incident on the test lens and an exit window for observing the image formation position of the test lens, a chart image of the image formation position is observed from outside the exit window of the thermostatic chamber with an objective lens having a long working distance, the working distance WD of the objective lens used here satisfies 10 mm or more, and spherical aberration due to the exit window is corrected.

3. The flange back variation measuring method or measuring device thereof according to claim 1, characterized in that the two or more resolution charts use a linear slit or a cross slit having a line width of 5 μm to 50 μm, the resolution chart used in the means for detecting the best imaging position of the first test lens is formed as a negative type slit whose slit portion is transparent, and the resolution chart used in the means for detecting the second chart surface position is formed as a positive type slit whose slit portion is light-shielded.

4. The method and device for measuring flange back variation according to claim 1 or 2, characterized in that a lens support that is placed in contact with the flange surface of the test lens and a member for mounting a resolution chart used in the means for detecting the second chart surface position are made of the same material and directly connected to each other.

5. The lens holder and the member for mounting the resolution chart used in the second position detection means have a linear expansion coefficient of 3.0×10 －６ 5. The flange back fluctuation measuring method or measuring device according to claim 4, wherein the flange back fluctuation is equal to or smaller than 1 / K.

6. The flange back variation measuring method or measuring device according to any one of claims 1 to 5, characterized in that a true flange back temperature variation ΔFBr is obtained by correcting an expansion / contraction component ΔL of the flange back temperature variation of a distance L from the flange surface of the test lens to the position of the second chart surface position due to temperature variation. When the distance from the flange surface to the second chart surface position is L and the linear expansion coefficient of the lens stand is α, the expansion / contraction correction amount ΔL of the lens stand is calculated as follows. (Math. 1) ΔL=L×α×ΔT Next, when the distance between the image forming point of the first chart and the surface of the second chart is H and the temperature variation of the distance H is ΔH, the true flange back temperature variation ΔFBr can be obtained from the following: (Math. 2) ΔFBr=ΔH−ΔL

7. The method for measuring flange back fluctuation according to claim 1 or 2, or the measuring device thereof, characterized in that the first position detection means and the second position detection means perform position detection from through-focus characteristic data of a modulation transfer function (MTF value) by using at least two or more processing functions among the following data processing: Frequency selection means: The function to select the MTF evaluation frequency to obtain a sharp and sharp MTF through-focus characteristic. ・Noise filtering: A function to reduce noise components of MTF through-focus characteristics using Low Pass Filter, Band Pass Filter, High Pass Filter, etc. - Approximate curve fitting process: A function that reduces detection errors caused by noise by fitting the MTF through-focus characteristics with a least squares function of a higher order function or a spline function. ・Threshold median calculation: A function that sets a threshold value for the MTF value and detects the best focus position by using the median value of the intercept at that time.

Images