WO2017020829A1 - 解像力测试方法和解像力测试装置 - Google Patents

解像力测试方法和解像力测试装置 Download PDF

Info

Publication number
WO2017020829A1
WO2017020829A1 PCT/CN2016/093018 CN2016093018W WO2017020829A1 WO 2017020829 A1 WO2017020829 A1 WO 2017020829A1 CN 2016093018 W CN2016093018 W CN 2016093018W WO 2017020829 A1 WO2017020829 A1 WO 2017020829A1
Authority
WO
WIPO (PCT)
Prior art keywords
line
image
resolution
wedge
brightness
Prior art date
Application number
PCT/CN2016/093018
Other languages
English (en)
French (fr)
Inventor
丁亮
廖海龙
王明珠
张宝忠
Original Assignee
宁波舜宇光电信息有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 宁波舜宇光电信息有限公司 filed Critical 宁波舜宇光电信息有限公司
Publication of WO2017020829A1 publication Critical patent/WO2017020829A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M11/00—Testing of optical apparatus; Testing structures by optical methods not otherwise provided for

Definitions

  • the present invention relates to the field of image processing, and more particularly to a resolution test method and a resolution test apparatus for testing the resolution of an image.
  • the resolution is used to describe the physical quantity of the microfilming system to reproduce the subtle parts of the original. It is an important indicator of image quality evaluation. In the field of camera and photography, it refers to the ability of the camera to express details. The image taken by a camera with good resolution is definitely The clarity is unmistakable. Conversely, images taken with poor resolution are prone to losing many details visible to the naked eye.
  • the camera module needs to test the resolution of the lens to accurately judge the resolution of the lens, and ensure that the lens of the factory can meet the application requirements.
  • Tvline TV scan line that can be output between computers and cameras
  • the ISO12233 standard card and HYRes analysis software are used for analysis to obtain the resolution of the lens.
  • Figure 7 is a schematic diagram of an ISO resolution chart.
  • the ISO resolution chart contains various patterns, mainly using J1 and K1 in the horizontal direction, J2 and K2 in the vertical direction, and JD and KD in the 45-degree direction.
  • the pattern in Fig. 7 is mainly represented by a wedge line pattern in which the number of patterns ⁇ 100 is the number of lines per unit height in the picture as a value for evaluating the resolution.
  • the chart is not necessarily used directly, and the desired portion, such as one or more wedge-shaped line patterns shown in FIG. 7, may be cut out from the chart and used after being re-spliced.
  • the captured image is printed or displayed, and the image is visually evaluated to evaluate the resolution of the camera.
  • the evaluation includes visual evaluation and evaluation using computer software, such as the above-mentioned HYRes analysis software. .
  • ISO-recommended software such as HYRes analysis software developed by Olympus
  • the test efficiency is not high, which is not conducive to the manufacturer's mass production on the production line.
  • Another object of the present invention is to provide a resolution test method and a resolution test device which improve algorithm stability and operational efficiency, and have high test stability and operability.
  • Another object of the present invention is to provide a resolution test method and a resolution test apparatus which are capable of adaptively adjusting a deflection angle without requiring an image to be rotated, which is highly operability and saves time.
  • Another object of the present invention is to provide a resolution test method and a resolution test apparatus which employ a targeted local denoising method, which avoids the need for more time for denoising and greatly improves the operation efficiency.
  • Another object of the present invention is to provide a resolution test method and a resolution test device that employ adaptive image enhancement technology to reduce the influence of external light source brightness on test results.
  • Another object of the present invention is to provide a resolution test method and a resolution test device, which adopts a method of coarse detection and fine inspection in performing resolution scan calculation, thereby realizing fast and accurate positioning of resolution positions and improving the algorithm. operating efficiency.
  • Another object of the present invention is to provide a resolution test method and a resolution test apparatus capable of stably and accurately reflecting the lens resolution in real time and quickly calculating the corresponding Tvline line logarithm.
  • Another object of the present invention is to provide a resolution test method and a resolution test device which are simple in operation, convenient to use, and suitable for popularization and application.
  • the present invention provides a resolution test method suitable for testing the resolution of a lens, wherein the method comprises the following steps:
  • step (C) only the test area is denoised.
  • the step (B) comprises the steps of: (B.1) automatically positioning a central position of a plurality of marking points on a label; (B.2) calculating an angle between the marking plate and the horizontal direction, determining a deflection angle between the plate and a module; and (B.3) a resolution scan along the deflection angle.
  • the resolution test method further includes a step (E) of detecting image brightness and comparing it with a standard brightness, wherein the step (E) is located before the step (D) and is located after the step (B) When the detected overall brightness of the image is lower than the standard brightness, the image is enhanced.
  • the image is enhanced by using an adaptive image enhancement technique.
  • the step (D) comprises the steps of: (D.1) segmenting the image, detecting the start line and the end line of the wedge line of the test area; (D.2) rough inspection; (D.3) fine Check; and (D.4) calculate the line logarithm.
  • the coarse inspection step is adapted to quickly locate to the vicinity of the resolvable position
  • the fine inspection step is provided for progressive scanning, precise positioning To the final resolution position.
  • a denoising method suitable for use in a resolution test process wherein the denoising method is local denoising.
  • An image brightness enhancement method wherein the method comprises the steps of: (1) detecting image brightness; (2) comparing with a standard brightness; and (3) comparing image brightness when the image brightness is lower than the standard brightness Enhance processing.
  • the step (3) is not required to be performed.
  • an adaptive image enhancement technique is employed.
  • the standard brightness is a preset brightness, which is pre-positioned in an image acquisition system.
  • the standard brightness is brightness suitable for testing.
  • the above resolution test method and the resolution test device are simple and convenient, have high operability, high precision, and high test efficiency, so that image deflection, noise and ambient brightness have little influence on the test result, and the resolution of the lens can be reflected in real time and accurately. So that it can be used in batches on the production line.
  • FIG. 1 is a block diagram of a test system in accordance with a preferred embodiment of the present invention.
  • FIG. 2 is a schematic illustration of an automatic search for a test area in accordance with a preferred embodiment of the present invention.
  • 3A and 3B are diagrams of an algorithm employed in accordance with a preferred embodiment of the present invention with Olympus A comparison of test results of HYRes analysis software.
  • FIG. 4 is a schematic diagram of a test flow in accordance with a preferred embodiment of the present invention.
  • Figure 5 is a schematic illustration of a test flow in accordance with a preferred embodiment of the present invention.
  • Figure 6 is a schematic illustration of enhanced image brightness in accordance with a preferred embodiment of the present invention.
  • Figure 7 is a schematic diagram of an ISO resolution chart.
  • a resolution testing system in accordance with a preferred embodiment of the present invention generally includes a standard 10, a light source 20, a module 30, an image acquisition system 40, and test software 50, wherein the positioning is performed at an appropriate position.
  • the stencil 10 and the light source 20 are then obtained by the module 30 and the image acquisition system 40, and then analyzed and tested by the test software 50, and finally the resolution value is quickly and accurately calculated.
  • the image acquisition system 40 is electrically coupled to the test software 50.
  • the resolution test method mainly comprises: acquiring at least one image via the test lens; identifying at least one test area in the image; and calculating the resolution value using a predetermined algorithm.
  • the test area in the image is a wedge line pattern as shown in FIG. 7, and the algorithm for calculating the resolution value is an algorithm for simulating a manual visual resolution of the visual resolution, for example, visually measuring with a measurer.
  • the manual visual evaluation only two conditions are agreed, that is, a) the spatial resolution of the number of wedge lines of the visual resolution evaluation pattern is changed (eg, from 5 to 4) as the resolution. , and b) must be observed from the low frequency side when observing. Therefore, through the above constraints, the calculation result by the software algorithm has good consistency with the manually visualized test value.
  • the traditional method is to press the entire image. Rotate at this deflection angle to eliminate the effect of rotation on the calculation results.
  • some interpolation including pre-interpolation
  • the interpolation method also affects the resolution, which seriously affects the operation efficiency.
  • the whole image is denoised, but in the actual calculation, the entire image is not used. For example, only the wedge line pattern used for the test needs to be identified and the resolution is calculated. This overall denoising approach results in lower computational efficiency.
  • the contrast of the black and white lines in the wedge line pattern is greatly affected by the brightness of the light source, resulting in a situation in which the conventional test method cannot be tested under a certain brightness condition.
  • the light source needs to be in accordance with ISO7589, using "daylight” (standard setting) or "tungsten lamp", and the illumination difference between any part of the chart and the central area is within ⁇ 10%. Inside. Therefore, since the brightness is not particularly required, the above problem may be encountered.
  • the testing method provided by the present invention includes the following steps:
  • step (a) by illuminating the module 30 to obtain a current image, it may be implemented to acquire an image by photographing, and extract a luminance Y component of the current image.
  • the step (b) includes the following steps: (b.1) automatically positioning a central position of the plurality of marking points on the label 10; (b.2) determining the marking 10 and the module 30 The angle of deflection between; and (b.3) adaptively adjusts the angle of deflection.
  • the center position of the plurality of marking points on the label 10 is automatically positioned by the testing software 50, preferably, the label as shown in FIG. Version 10, using image processing technology, automatically locates the center positions (O 1 , O 2 , O 3 , O 4 ) of several marked points, such as 4 Mark points in the preferred embodiment.
  • the center positions (O 1 , O 2 , O 3 , O 4 ) of several marked points such as 4 Mark points in the preferred embodiment.
  • it is not limited to black dots, but may be other feature shapes such as squares and triangles.
  • the center coordinates of the test area can be determined by the center coordinates of the image and W and H.
  • the center coordinate of each test area can be determined by the ratio of the distance between the center of the test area and the center of the image in the horizontal and vertical directions to W/2 and H/2.
  • the angle of the deflection between the plate 10 and the module 30 is conveniently determined by calculating the angle between the plate 10 and the horizontal direction.
  • the adaptive adjustment of the deflection angle is performed by adopting a resolution scan along the deflection angle.
  • the resolution test method of the embodiment of the invention proposes a scheme for adaptively adjusting the deflection angle without loss of efficiency.
  • the angle ⁇ between the standard 10 and the horizontal direction can be calculated.
  • ⁇ is non-zero
  • the standard is indicated.
  • the conventional method is to rotate the entire image according to this angle to eliminate the influence of the rotation on the calculation result, so that the influence of the deflection angle on the test result can be well eliminated.
  • the operation of rotating images is complicated, it takes a lot of time and seriously affects the calculation efficiency. If used to batch verify the resolution of the camera module, the overall work efficiency will be much lower.
  • the resolution test method of the embodiment of the present invention adopts a scheme of performing resolution scanning along the deflection angle, so that it is not necessary to rotate the image, and the angle of deflection can be accommodated as well, without any influence on efficiency.
  • step (c) a targeted denoising method is adopted to improve the stability and operating efficiency of the algorithm.
  • the general rule is that the better the denoising effect is, the lower the computational efficiency is. This is because the denoising process takes time and the efficiency of the operation is reduced invisibly.
  • the resolution test method of the embodiment of the present invention proposes a targeted denoising method.
  • a complete image not all places are used, and may only be used.
  • the resolution test method according to the embodiment of the present invention only the calculation area or the test area is denoised, and no untested area, such as a blank area other than the wedge line pattern, is denoised, thereby saving noise reduction. Time, improve the efficiency of denoising, and thus greatly improve the efficiency of computing.
  • the denoising time of the entire image is about 2 s for an image of 8 M in size
  • the local denoising scheme provided by the embodiment of the present invention is used only for the area to be calculated.
  • the denoising time can be controlled within 50ms, and when used in batches on the production line, the local, targeted denoising scheme provided by the embodiment of the present invention, and the conventional Compared with the overall denoising scheme, the efficiency of denoising can be greatly improved.
  • the image acquisition system 40 and the test software 50 can automatically analyze whether the collected image has a noise condition, and if there is noise, can automatically determine the degree of influence of the noise on the analysis test, if any, The image is automatically denoised, and if it is not affected, no denoising is required.
  • the step (d) includes the steps of: (d.1) detecting image brightness; (d.2) comparing with a standard brightness; and (d.3) when the image brightness is lower than the Enhance image brightness when using standard brightness.
  • the light source 20 captures a test image whose brightness meets the requirements, and may require repeated tests. By repeatedly adjusting the position and brightness of the light source 20, an image that meets the requirements can be captured, and the dimming and shooting are repeated. During the test, the test efficiency will be seriously reduced.
  • an adaptive image enhancement technique is applied in the image acquisition system 40 to automatically detect the overall brightness of the image.
  • a standard brightness is input, and the detected image brightness is compared with the standard brightness.
  • the image is automatically enhanced to achieve the standard brightness.
  • the standard brightness is the brightness required by the test software provided by the embodiment of the present invention. After the standard brightness is reached, the test can be performed, and at the same time, the test result is relatively accurate, wherein the standard brightness can be Set to a value or an interval.
  • FIG. 3A shows the results of testing using Olympus HYRes software in the case where the overall brightness of the image is 130
  • FIG. 3B shows the embodiment of the present invention in the case where the overall brightness of the image is 130.
  • the test results of the resolution test method provided. It can be seen from the comparison between FIG. 3A and FIG. 3B that when the overall brightness of the image is 130, the Olympus HYRes software cannot perform the test, and the brightness does not reach the brightness required by the HYRes software test, and the self-provided by the embodiment of the present invention.
  • Adapted to image enhancement technology when the brightness of the image is not met, the image brightness is automatically enhanced to enhance the brightness of the image to meet the test requirements, and then tested by HYRes software.
  • the algorithm of the embodiment of the present invention can automatically adjust the brightness of the image. Therefore, the brightness of the external light source is less restricted, the utility is stronger, the brightness of the external light source is lower, and the application range is wider. Moreover, the steps of repeatedly adjusting the light source and repeatedly acquiring images are not required, thereby improving the test efficiency and being suitable for batch application on the production line.
  • step (c) may be performed before the step (d) or after the step (d), and the step (c) and the step (d) may also be performed. At the same time. That is to say, denoising and enhancing image brightness can be performed separately or simultaneously.
  • step (e) when the resolution scan calculation is performed, the method of the coarse inspection and the fine inspection is adopted, thereby realizing the rapid positioning resolution position and improving the operation efficiency of the algorithm.
  • the step (e) includes the following steps: (e.1) segmenting the image, detecting a start line and an end line of the wedge line of the test area; (e.2) rough inspection; (e.3) Fine inspection; and (e.4) calculate the line logarithm.
  • the image is segmented, and then the start line and the end line of the wedge line are detected, that is, the wedge line to be detected is determined, so as to facilitate subsequent rapid coarse inspection and Fine inspection makes the positioning of rough inspection and fine inspection more rapid and accurate.
  • the method of the first coarse inspection and the fine inspection proposed by the invention can effectively scan the efficiency.
  • the step (e.2) by the coarse inspection, the vicinity of the resolvable position can be quickly located, and then the step (e.3) is performed, and then the fine inspection is performed line by line, and the precise positioning is performed.
  • the final resolution position which can eliminate the step of scanning the unresolved position, can effectively improve the scanning efficiency, and then quickly calculate the Tvline line number, calculate the resolution value, and accurately reflect the resolution of the lens.
  • the coarse inspection step length x needs to satisfy the following formula:
  • the method of the first rough detection and fine inspection provided by the present invention is much more efficient than the conventional progressive scan resolution method.
  • the wedge start line is detected in the identified test area, the main scanning direction of the image crosses the wedge line, and the image data is horizontally scanned from the wider end of the wedge line, and the Wedge start line WSL is detected. For example, every Reading the 1 line selects the smallest 3 values. When the difference between the average of the 3 values and the line full point average is 5 times of the initial line, it is interpreted as the beginning of the wedge shape.
  • the scanning is continued, and the maximum value and the minimum value of each line are detected in the count of the number of wedge lines.
  • the change of the threshold is ignored.
  • the threshold moves toward the high frequency, the threshold is gradually reduced and the detection is repeated, so that the minimum and maximum values can be counted.
  • the line is taken as the resolution limit line LML.
  • Scanning is then performed to find the final line WEL of the wedge line pattern. For example, the amplitude of each line can be obtained and the amplitude of the previous line can be continuously compared, and when the sharp decrease occurs, the final line can be determined.
  • the image is first scanned in step size x to obtain the value of the resolution limit line LML, but this can only explain that the detected resolution limit line LML is in the step size.
  • the range it is not possible to accurately determine which line the resolution limit line is in. Therefore, it is possible to further perform a progressive scan from the range of -x to +x of the resolution limit line to determine an accurate resolution.
  • the value of the limit line LML is reduced from the height range of the entire wedge line pattern in the original test area to the range of twice the step size, so that the amount of calculation is greatly reduced.
  • the image in order to accurately detect the start line and the end line of the wedge line, when the image is scanned in the step size x, it is preferably also at the start line and the end line of the wedge line A progressive scan is further performed in the range of -x to +x to accurately determine the values of the start and end lines of the wedge line.
  • the image may be scanned line by line first, and then scanned in steps x after the start line of the wedge line is detected.
  • the image may be scanned progressively after detecting the resolution limit line in step x to detect the end line of the wedge line.
  • step size x when the value of the step size x is large, the calculation amount required for the coarse inspection step is small, but the calculation amount required for the fine step is large, and when the value of the step size x is larger than In hours, the amount of calculation required for the rough inspection step is large, but the amount of calculation required for the detailed step is small, so the step size of the rough inspection step can be flexibly set according to the specific situation, and only the above relationship needs to be satisfied.
  • the resolution test method according to the embodiment of the present invention is simple and easy to operate, and has high computational efficiency.
  • the algorithm is accurate, the test result is accurate, and it can better reflect the resolution of the lens. It is suitable for batch use on the production line, which provides a simple and accurate method for the resolution test of the lens.
  • a resolution testing device which is coupled to a lens module to test the resolution of the lens module, wherein the resolution testing device includes:
  • An image obtaining unit configured to acquire at least one image by using the lens module
  • a region extracting unit configured to identify at least one test area from the image acquired by the image acquiring unit
  • a resolution calculation unit configured to calculate a resolution value in the test area extracted by the area extraction unit.
  • the region extracting unit is specifically configured to locate the center positions O 1 , O 2 , O 3 , O 4 of the four sides of the image; and the difference between the abscissas of the two points in the horizontal direction.
  • the ratio of the distance in the horizontal and vertical directions to the ratio of W/2 and H/2 determines the center coordinate of each test area.
  • the denoising unit performs denoising only on the test area extracted by the area extracting unit.
  • the test area includes a wedge line pattern
  • the resolution calculation unit is configured to calculate a resolution value by scanning the wedge line pattern.
  • the area extracting unit specifically includes: a positioning module for automatically positioning a center position of the plurality of marking points in the image; and an extracting module configured to connect the plurality of markers according to the center of the wedge line pattern The proportion of the multiple line segments of the point is different, and the wedge line pattern is automatically recognized.
  • the resolution testing device further comprising an angle correcting unit for calculating an angle between the image and the horizontal direction, and in the case where the angle is not zero, as a deflection angle between the image and the lens module
  • the resolution calculation unit is configured to scan the wedge line pattern along the deflection angle to calculate a resolution value.
  • the above-mentioned resolution testing device further includes a brightness adjusting unit, specifically comprising: a brightness detecting module for detecting an overall brightness of the image; a brightness comparing module for comparing the detected brightness with a standard brightness; and brightness enhancement A module for enhancing an image if the detected brightness is less than the standard brightness.
  • a brightness adjusting unit specifically comprising: a brightness detecting module for detecting an overall brightness of the image; a brightness comparing module for comparing the detected brightness with a standard brightness; and brightness enhancement A module for enhancing an image if the detected brightness is less than the standard brightness.
  • the brightness detection module detects the overall brightness of the image by acquiring the luminance Y component of the image.
  • the brightness enhancement module performs an enhancement process on the image using an adaptive image enhancement technique.
  • the resolution calculation unit scans the wedge line along the deflection angle calculated by the angle correction unit for the image denoised by the denoising unit and subjected to the enhancement processing by the brightness adjustment unit Pattern to calculate the resolution value.
  • At least two of the denoising unit, the angle correcting unit, and the brightness adjusting unit may operate in parallel, or the denoising unit, the angle correcting unit, and the brightness adjusting unit may be in any order. This operation.
  • the resolution calculating unit scans the wedge line pattern in a direction perpendicular to a line in the wedge line pattern from a wider end of the wedge line pattern to start a line and a wedge shape through the wedge line.
  • the line resolution limit line and the wedge line end line are used to calculate the resolution value.
  • the resolution calculation unit starts from the wider end of the wedge line pattern, and scans the wedge line pattern in a direction perpendicular to the line in the wedge line pattern; by scanning in each line of the scan
  • the gray level of the pixel determines the number of wedge lines included in the line; and in the case where the number of wedge lines determined in the current line is smaller than the number of wedge lines determined in the previous line, the current behavior of the resolution limit line is determined.
  • the resolution calculation unit is specifically configured to: scan the image by the step size x to detect the initial wedge line start line, the initial wedge line resolution limit line, and the initial wedge line end line; respectively, at the initial wedge line start line
  • the initial wedge line resolution limit line and the initial wedge line end line are scanned line by line from -x to +x to detect the final wedge line start line, the final wedge line resolution limit line, and the final wedge line end line.
  • the resolution calculation unit is specifically configured to: scan the image progressively to detect the final wedge line start line; scan the image in step size x to detect the initial wedge line resolution limit line; at the initial wedge line resolution limit Line-by-row scans are performed from -x to +x of the line to detect the final wedge line resolution limit line.
  • the resolution calculation unit is specifically configured to: scan the image by the step size x to detect the initial wedge line resolution limit line; scan the line by line in the range of -x to +x of the initial wedge line resolution limit line, The final wedge line resolution limit line is detected; the image is scanned progressively to detect the end line of the final wedge line.
  • the starting line of the wedge line pattern can also be identified by other methods, such as a computer artificial vision algorithm.
  • the limit and end lines are resolved to calculate the resolution value.
  • the resolution test apparatus includes the image acquisition system 40 and the test software 50 as shown in FIG. 1, wherein the image acquisition unit is embodied as the image acquisition system 40, and other units can The specific performance is in the form of test software 50.
  • the test software 50 can be machine code, firmware, embedded code, and application software.
  • the resolution test apparatus according to an embodiment of the present invention may also be implemented as hardware.
  • the hardware may be a circuit, a processor, a computer, an integrated circuit, an integrated circuit core, a microelectromechanical system (MEMS), a passive device, or a combination thereof.
  • MEMS microelectromechanical system
  • test software 50 may be embodied as computer readable code on a computer readable recording medium.
  • a computer readable recording medium is any data storage device that can store data that can thereafter be read by a computer system. Examples of the computer readable recording medium include read only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, and carrier wave (such as data transmission via the Internet).
  • the computer readable recording medium can also be distributed via a network coupled computer system such that the computer readable code is stored and executed in a distributed fashion.
  • ROM erasable or rewritable read only memory
  • RAM random access memory
  • IC IC
  • optically or magnetically recordable non-transitory machine readable for example, a computer readable storage medium, such as a compact disc (CD), a digital versatile disc (DVD), a magnetic disk, or a magnetic tape.
  • Methods and apparatus in accordance with embodiments of the present invention may be implemented by a computer or mobile terminal including a controller and memory, and the memory may be one or more programs adapted to store instructions including various embodiments for implementing the present invention.
  • a non-transitory machine readable for example, computer readable storage medium.
  • the resolution test apparatus may be electrically connected to the light source 20 illustrated in FIG. 1, and the brightness adjustment unit may enhance the brightness of the image by adjusting the brightness of the light source 20.
  • a resolution test method for testing the resolution of a lens module includes:
  • the resolution value in the extracted test area is calculated.
  • the method further includes: denoising the acquired image.
  • only the extracted test area is denoised.
  • the test area includes a wedge line pattern
  • calculating the extracted resolution value in the test area is specifically calculating the extracted resolution in the test area by scanning the wedge line pattern. value.
  • identifying at least one test area from the acquired image specifically includes automatically positioning a center position of a plurality of mark points in the image; and connecting the plurality of mark points according to a center of the wedge line pattern The proportion of the multiple line segments is different, and the wedge line pattern is automatically recognized.
  • the method further includes: calculating an angle between the image and the horizontal direction, and the angle is not zero. In the case, as a deflection angle between the image and the lens module; and calculating the extracted resolution value in the test area, specifically scanning the wedge line pattern along the deflection angle to calculate a resolution value.
  • the image is enhanced.
  • the overall brightness of the image is detected by acquiring the luminance Y component of the image.
  • the image is enhanced using adaptive image enhancement techniques.
  • At least two of the denoising step, the angle correcting step, and the brightness adjusting step may be performed in parallel, or the denoising step, the angle correcting step, and the brightness adjusting step may be performed in any order This is done.
  • calculating the resolution value in the extracted test region is specifically: starting from a wider segment of the wedge line pattern, scanning the wedge line pattern in a direction perpendicular to a line in the wedge line pattern The resolution value is calculated by the wedge line start line, the wedge line resolution limit line, and the wedge line end line.
  • the resolution calculating step starts from a wider end of the wedge line pattern to be perpendicular to the wedge
  • the direction of the line in the line pattern scans the wedge line pattern; the number of wedge lines included in the line is determined by the gray level of the scanned pixel in each line of the scan; and in the current line
  • the current behavior of the resolution limit line is determined.
  • calculating the extracted resolution value in the test area is specifically: scanning the image in step size x to detect an initial wedge line start line, an initial wedge line resolution limit line, and an initial wedge line end line;
  • the initial wedge line start line, the initial wedge line resolution limit line, and the initial wedge line end line are scanned line by line from -x to +x to detect the final wedge line start line, the final wedge line resolution limit line, and finally The wedge line ends.
  • calculating the extracted resolution value in the test area is specifically: a progressive scan image to detect a final wedge line start line; scanning the image in a step size x to detect an initial wedge line resolution limit line; The line-by-line scan is performed in the range of -x to +x of the initial wedge line resolution limit line to detect the final wedge line resolution limit line.
  • calculating the extracted resolution value in the test area is specifically: scanning the image in step size x to detect an initial wedge line resolution limit line; in the range of -x to +x of the initial wedge line resolution limit line In-line scan to detect the final wedge line resolution limit line; progressively scan the image to detect the end of the final wedge line.
  • Still another preferred embodiment of the present invention provides a denoising method suitable for use in a resolution test process, wherein the denoising method is local denoising.
  • a further preferred embodiment of the present invention provides an image brightness enhancement method, wherein the method comprises the steps of: detecting image brightness; comparing with a standard brightness; and in the case where the image brightness is lower than the standard brightness, the image is Brightness is enhanced.
  • the image brightness is not enhanced.
  • an adaptive image enhancement technique is employed.
  • the standard brightness is a preset brightness, which is pre-positioned in an image acquisition system.
  • the standard brightness is brightness suitable for testing.

Landscapes

  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Testing Of Optical Devices Or Fibers (AREA)
  • Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)

Abstract

一种解像力测试方法，包括以下步骤：（A）获取至少一图像；（B）识别出至少一测试区域；（C）去噪；以及（D）计算分辨率值。该解像力测试方法通过自适应调整旋转角度、自适应图像增强处理及有针对性的局部去噪，简单易操作，并具有较高的运算效率，适于在生产线上批量使用。

Description

解像力测试方法和解像力测试装置 技术领域
本发明涉及图像处理领域，尤其涉及用于测试图像的解像力的解像力测试方法和解像力测试装置。
背景技术
解像力是用来描述缩微摄影系统再现被摄原件细微部分能力的物理量，是影像量评价的重要指标，在摄像、照相领域，是指相机对细节的表现能力，解像力好的相机拍摄的图像肯定是毫发毕现般的清晰，反之，解像力较差的拍摄出来的图像则容易丢失许多肉眼可见的细节。
因此，摄像模组在生产过程中需要测试镜头的解像力，以准确地判断其解像力情况，保证出厂的镜头能够满足应用需求。目前，ISO提出的Tvline(电脑、相机之间能输出的电视扫描线)测试方法被很多厂家认可，使用较多的是ISO12233标准卡结合HYRes分析软件进行分析来得到镜头的解像力。
图7是ISO分辨率图表的示意图。如图7所示，ISO分辨率图表中包含各种样式，主要使用的是其中的水平方向的J1、K1；垂直方向的J2、K2和倾斜45度方向的JD、KD。图7中的样式主要表现为楔形线图案，其中图案的数字×100就是画面中每单位高度的条数，以作为评价解像力的数值。
在实际使用中，并不一定直接使用该图表，也可以从该图表中裁剪出需要的部分，例如图7所示的一个或多个楔形线图案，并经过重新拼接排列后使用。此外，也可以自己制作并使用与图7所示的ISO分辨率图表相同的图表，但是需要满足ISO12333中规定的具体事项，例如白底部分的反射率与黑色部分的反射率之比、各个图案的位置精度等。
在对上述ISO分辨率图表进行拍摄之后，打印或者显示所拍摄的图像，并对该图像进行视觉评估来评估相机的解像力，评估包括目测评估和利用计算机软件来进行评估，比如上述的HYRes分析软件。
但是ISO推荐的软件，比如奥林巴斯开发的HYRes分析软件在计算时，测 试效率不高，从而不利于制造厂家在生产线上批量使用。
发明内容
本发明的一个目的在于提供解像力测试方法和解像力测试装置，其测试效率较高，适于在生产线上批量使用。
本发明的另一目的在于提供解像力测试方法和解像力测试装置，其提高了算法稳定性和运行效率，具有较高的测试稳定性和可操作性。
本发明的另一目的在于提供解像力测试方法和解像力测试装置，其能够自适应调整偏转角度，不需要旋转图像，操作性强，节约时间。
本发明的另一目的在于提供解像力测试方法和解像力测试装置，其采用了有针对性的局部去噪方法，避免了去噪花费较多的时间，大大提高了运算效率。
本发明的另一目的在于提供解像力测试方法和解像力测试装置，其采用了自适应图像增强技术，减少了外界光源亮度对测试结果的影响。
本发明的另一目的在于提供解像力测试方法和解像力测试装置，其在进行分辨率扫描计算时，采用了先粗检后细检的方法，实现了快速、精准地定位分辨率位置，提高了算法运行效率。
本发明的另一目的在于提供解像力测试方法和解像力测试装置，其能够实时稳定且准确地反映出镜头解像力，并快速的计算出相应的Tvline线对数。
本发明的另一目的在于提供解像力测试方法和解像力测试装置，其操作简单，使用方便，适于推广应用。
为满足本发明的以上至少一个发明目的，本发明提供一种解像力测试方法，适于测试镜头的解像力，其中所述方法包括以下步骤：
(A)获取至少一图像；
(B)识别出至少一测试区域；
(C)去噪；以及
(D)计算分辨率值。
其中在所述步骤(C)中，只对所述测试区域进行去噪。
所述步骤(B)包括以下步骤：(B.1)自动定位出一标版上的多个标记点的中心位置；(B.2)计算出所述标版与水平方向的夹角，确定所述标版与一模组之间的偏转角度；以及(B.3)沿着所述偏转角度进行分辨率扫描。
所述解像力测试方法进一步包括一步骤(E)：检测图像亮度，并与一标准亮度进行对比，其中所述步骤(E)位于所述步骤(D)之前，并位于所述步骤(B)之后，当检测到的图像整体亮度低于所述标准亮度时，则对图像进行增强处理。
其中在所述步骤(E)中，采用自适应图像增强技术，对图像进行增强处理。
所述步骤(D)包括以下步骤：(D.1)对图像进行分割，检测所述测试区域的楔形线的起始行和结束行；(D.2)粗检；(D.3)细检；以及(D.4)计算出线对数。
其中在所述步骤(D.2)和所述步骤(D.3)中，所述粗检步骤适于快速定位到可分辨位置的附近，提供所述细检步骤进行逐行扫描，精确定位到最终的分辨率位置。
优选地，所述粗检步长需要满足如下公式：F(x)＝2x2-ax+a(a为区域扫描范围)。
一种去噪方法，适于应用于解像力测试过程中，其中所述去噪方法为局部去噪。
其中去噪时，只对计算区域进行去噪。
一种图像亮度增强方法，其中所述方法包括以下步骤：(1)检测图像亮度；(2)与一标准亮度进行对比；以及(3)当图像亮度低于所述标准亮度时，对图像亮度进行增强处理。
其中当检测到的图像亮度不低于所述标准亮度时，则不需执行所述步骤(3)。
其中在上述方法中，采用的是自适应图像增强技术。
优选地，所述标准亮度为预设亮度，预先置于一图像采集系统。
优选地，所述标准亮度为适于测试的亮度。
以上解像力测试方法和解像力测试装置简单方便，可操作性强，精准度高，测试效率也较高，使得图像偏转、噪声和环境亮度对测试结果影响较小，可以实时且准确的反应镜头的解像力，从而能够在生产线上批量使用。
附图说明
图1是根据本发明的一个优选实施例的测试系统的框图。
图2是根据本发明的一个优选实施例中自动寻找测试区域的示意图。
图3A和图3B是根据本发明的一个优选实施例采用的算法与奥林巴斯的 HYRes分析软件的测试结果对比示意图。
图4是根据本发明的一个优选实施例的测试流程示意图。
图5是根据本发明的一个优选实施例的测试流程示意图。
图6是根据本发明的一个优选实施例的增强图像亮度的示意图。
图7是ISO分辨率图表的示意图。
具体实施方式
以下描述用于揭露本发明以使本领域技术人员能够实现本发明。以下描述中的优选实施例只作为举例，本领域技术人员可以想到其他显而易见的变型。在以下描述中界定的本发明的基本原理可以应用于其他实施方案、变形方案、改进方案、等同方案以及没有背离本发明的精神和范围的其他技术方案。
图1是根据本发明的一个优选实施例的测试系统的框图。如图1所示，根据本发明实施例的解像力测试系统总的来说包括一标版10、一光源20、一模组30、一图像采集系统40和测试软件50，其中在适当的位置定位该标版10和该光源20，然后通过该模组30和该图像采集系统40获取一张当前图像，进而通过所述测试软件50进行分析测试及计算，最终快速准确地计算出分辨率值，以精准地反映镜头的解像力，其中所述图像采集系统40电连接于所述测试软件50。
根据本发明的优选实施例，解像力测试方法主要包括：经由测试用镜头获取至少一图像；识别出图像中的至少一测试区域；和，利用预定算法计算分辨率值。
在上述解像力测试方法中，图像中的测试区域是如图7所示的楔形线图案，且用于计算分辨率值的算法是用于模拟人工目测求解视觉分辨率的算法，例如与测量者目测判定分辨率极限点几乎相同的判定算法。具体地说，在人工目测评估时，仅约定了两个条件，即，a)将视觉分辨率评估图案的楔形线数发生变化(如由5条变为4条)的空间分辨率作为分辨率，和b)观察时必须从低频侧开始。因此，通过上述约束条件，通过软件算法的计算结果与人工目测的测试值具有良好的一致性。
但是，在利用算法来计算分辨率值的解像力测试方法中，受图像旋转、噪声、环境亮度等因素的影响很大，使得测试结果的重复性很差，不利于制造厂家在生产线上批量使用。
例如，当标版和镜头模组之间存在偏转角度时，传统的方法是将整张图像按 照这个偏转角度进行旋转，以消除旋转对计算结果的影响。但是，由于通过算法旋转图像，尤其是旋转非90度的角度时必须进行某种插值(包括前置插值)，同时插值方法也会对分辨率造成影响，这种方式会严重影响运算效率。
此外，在传统的方法中，采用的是对整幅图像进行去噪的方式，而实际计算中，并没有用到整幅图像，例如，仅需要识别出测试所用的楔形线图案并计算分辨率，这种整体去噪的方式导致运算效率较低。
另外，在传统的方法中，楔形线图案中的黑白线对比度受光源亮度影响较大，导致在某一亮度情况下，传统的测试方法会出现无法测试的情形。在ISO12233标准中，仅规定了光源需要按照ISO7589的规定，采用“日光”(标准设定)或者“钨丝灯”，且照明时保证图表任何部分与中央区域的照度差异位于±10％的范围内。因此，由于并没有对亮度进行特别要求，所以可能会遇到上述问题。
如图4和图5所示，本发明提供的测试方法包括以下步骤：
(a)获取至少一张当前图像；
(b)识别出至少一测试区域；
(c)去噪；
(d)检测图像亮度；以及
(e)计算分辨率值。
其中，在所述步骤(a)中，通过点亮所述模组30，获取一张当前图像，可以实施为通过拍照的方式获取图像，并提取当前图像的亮度Y分量。
当然，本领域技术人员可以理解，如果不对图像的亮度进行处理，则不需要提取图像的亮度Y分量。
所述步骤(b)包括以下步骤：(b.1)自动定位出所述标版10上的多个标记点的中心位置；(b.2)确定所述标版10与所述模组30之间的偏转角度；以及(b.3)自适应调整所述偏转角度。
具体地，在所述步骤(b.1)中，通过所述测试软件50自动定位出所述标版10上的多个标记点的中心位置，优选地，如图2所示的所述标版10，采用图像处理技术，先自动定位出若干个标记点的中心位置(O1，O2，O3，O4)，如本优选实施例中的4个Mark(标记点)黑圆点，但不仅限于黑圆点，也可以如方块、三角等其他特征形状。然后，将水平方向上两点的横坐标的差值定义为水平方向宽度，即W＝O3x-O1x，并垂直方向上两点的纵坐标的差值定义为垂直方向高度， 即H＝O4y-O2y。这样，测试区域的中心坐标就可以由图像的中心坐标以及W和H确定。具体地说，每个测试区域的中心坐标可以通过测试区域中心与图像中心在水平和垂直方向上的距离占W/2和H/2的比例确定。
在所述步骤(b.2)中，通过计算出所述标版10与水平方向的夹角，进而适于方便的确定所述标版10与所述模组30之间的偏转角度。在所述步骤(b.3)中，通过采用沿着所述偏转角度进行分辨率扫描的方式进行所述偏转角度的自适应调整。
这是由于在实际生产过程中，很容易出现所述30模组和所述标版10之间产生一定的偏转角度，但这个所述偏转角度对计算结果影响很大，为了解决此问题，本发明实施例的解像力测试方法提出了一种无效率损失的自适应调整偏转角度的方案。
如图2所示，通过定位出的所述标记点的中心位置O1和O3，可计算出所述标版10与水平方向的夹角θ，当θ非零时，说明所述标版10和所述模组30之间存在偏转角度，传统方法是将整张图像按照这个角度进行旋转，以消除旋转对计算结果的影响，这样虽然可以很好的消除偏转角度对测试结果的影响，但由于旋转图像的运算比较复杂，会花费较多的时间，并严重影响运算效率，如果用于批量验证摄像模组的解像力时，整个工作效率会降低很多。为了解决这个问题，本发明实施例的解像力测试方法采用沿着所述偏转角度进行分辨率扫描的方案，这样就不需要对图像进行旋转，同样可以适应偏转角度，而且对效率没有任何影响。
在所述步骤(c)中，采用了有针对性的去噪方法，提高了算法稳定性和运行效率。
在实际应用中，一般的规律是去噪效果越好的算法，运算效率越低，这是由于去噪的过程需要花费时间，无形中降低了运算的效率。
本发明实施例的解像力测试方法为了提高运算效率，提出了一种有针对性的去噪方法，由于在实际应用中，对于完整的一幅图像，并不是所有的地方都使用，可能只会用到需要计算的区域或者需要测试的区域，例如楔形线图案的区域，而传统的方法中，对整幅图像进行去噪，花费较多的时间。在根据本发明实施例的解像力测试方法中，只对计算区域或者测试区域进行去噪，不再对未测试的区域，例如楔形线图案之外的空白区域进行去噪，这样就节约了去噪时间，提高了去噪效率，进而大大提高了运算效率。
例如，通过对比试验发现，对于一幅大小为8M的图像，对整幅图像进行去噪的时间大约为2s，而采用本发明实施例提供的局部去噪的方案，即只对需要计算的区域或者需要识别的区域进行去噪，则去噪时间可以控制在50ms以内，当在生产线上批量使用的时候，采用本发明实施例提供的这种局部的、有针对性的去噪方案，和传统的整体去噪的方案相比，能够大大提高去噪运算效率。
值得一提的是，所述图像采集系统40和所述测试软件50能够自动分析采集的所述图像是否存在噪声情况，如果存在噪声，能够自动判断噪声对分析测试的影响程度，若有影响，则自动对所述图像进行去噪，如果不影响，则不需要去噪。
如图6所示，所述步骤(d)包括以下步骤：(d.1)检测图像亮度；(d.2)与一标准亮度进行对比；以及(d.3)当图像亮度低于所述标准亮度时，对图像亮度进行增强处理。
这是由于楔形线图案中的黑白线对比度受光源亮度影响较大，如果图像亮度较低，则会导致图像无法测试，或者导致测试结果偏差较大，为了不影响图像测试的准确性，需要调整所述光源20，拍摄出亮度符合要求的测试图像，有可能需要反复多次试验，通过反复调整所述光源20的位置及亮度等，才能拍摄出符合要求的图像，在反复调光及拍摄、测试的过程中，会严重的降低测试效率。
在根据本发明实施例的解像力测试方法中，为了减少外界光源亮度对测试结果的影响，提高测试效率，在所述图像采集系统40中应用自适应图像增强技术，自动检测图像的整体亮度，预先输入一所述标准亮度，并将检测到的图像亮度与所述标准亮度做对比，当图像整体亮度低于所述标准亮度时，自动对图像进行增强处理，使其达到所述标准亮度。值得一提的是，所述标准亮度是本发明实施例提供的测试软件所需的亮度，达到所述标准亮度后，才能够进行测试，同时，测试结果也较为准确，其中所述标准亮度可以设为一个值或者一个区间。
图3A所示为在图像整体亮度均为130的情况下，使用奥林巴斯的HYRes软件进行测试的结果，图3B所示为在图像整体亮度均为130的情况下，采用本发明实施例提供的解像力测试方法的测试结果。由图3A和图3B的对比可知，当图像整体亮度为130时，奥林巴斯的HYRes软件无法进行测试，亮度达不到HYRes软件测试所要求的亮度，而通过本发明实施例提供的自适应图像增强技术，当检测到图像亮度不符合要求时，自动对图像亮度进行增强处理，使图像亮度增强到符合测试要求的亮度的时候再进行测试，通过与HYRes软件的测试结 果相比，本发明实施例的算法由于能够自动调整图像亮度，因此，受外界光源亮度限制较少，实用性较强，对外界光源亮度的要求较低，应用范围较为广泛。而且不需要反复调整光源及反复获取图像的步骤，提高了测试效率，适于在生产线上批量应用。
值得一提的是，所述步骤(c)可以在所述步骤(d)之前进行，也可以在所述步骤(d)之后进行，所述步骤(c)及所述步骤(d)也可以同时进行。也就是说，去噪和增强图像亮度可以分开进行，也可以同时进行。
在所述步骤(e)中，在进行分辨率扫描计算时，采用了先粗检后细检的方法，实现了快速定位分辨率位置，提高了算法运行效率。
其中所述步骤(e)包括以下步骤：(e.1)对图像进行分割，检测所述测试区域的楔形线的起始行和结束行；(e.2)粗检；(e.3)细检；以及(e.4)计算出线对数。
其中在所述步骤(e.1)中，对图像进行分割，然后检测所述楔形线的起始行和结束行，即确定出待检测的所述楔形线，便于后续进行快速地粗检和细检，使得粗检和细检的定位更加迅速和准确。
在传统的测试中，即在标准的ISO12233的算法中，采用的是逐行扫描分辨率的方法，逐行扫描会花费较多的时间，使得测试效率非常低。
本发明提出的一种先粗检再细检的方法，可以有效扫描效率。其中在所述步骤(e.2)中，通过所述粗检，可以快速定位到可分辨位置的附近，然后执行所述步骤(e.3)，再逐行进行细检，精确的定位到最终的分辨率位置，这样可以省去扫描未分辨位置的步骤，因此可以有效提高扫描效率，进而快速的计算出Tvline的线对数，计算出分辨率值，得以精准的反应镜头的解像力。
值得一提的是，粗检时，粗检步长x需要满足如下公式：
F(x)＝2x2-ax+a(a为区域扫描范围)
优选地，当F(x)＜0时，本发明提供的先粗检再细检的方法比传统的逐行扫描分辨率的方法效率要高很多。
下面介绍根据本发明实施例的解像力测试方法中采用的计算分辨率值的一示例算法。
首先在识别出的测试区域中检测楔形开始线，图像的主扫描方向横切楔形线，从楔形线较宽的一端开始水平扫描图像数据，检测楔形开始行WSL。例如，每 读1条线选出最小的3个值，当3个值的平均值与线全点平均值之差为最初线的5倍时，即解释为开始出现楔形。
继续扫描，在楔形线数的计数中检测各线的极大值和极小值。此时为了去除波动的影响，忽略未满阈值的变化。阈值随着扫描线向高频移动，逐渐减小阈值并重复检测，这样即可数出极小值和极大值。随着操作的进行，当可数的黑线数与初始的楔形线数不一致时，将该线作为分辨率极限线LML。
接着进行扫描，可找到楔形线图案的最终线WEL。例如，可以通过求各线的振幅，并与前一条线的振幅继续比较，当出现急剧减少时即可判断为最终线。
最后，利用WSL、LML和WEL计算分辨率。例如，当楔形线的条数为5条时，分辨率为(100+500×(LML-WSL)/(WEL-WSL))×C，而当楔形线的条数为9条时，分辨率为(500+1500×(LML-WSL)/(WEL-WSL))×C，其中C为修正系数，通过整个画面高度PHT与楔形线长度(WEL-WSL)来确定，例如C＝0.3×PHT/(WEL-WSL)。
在根据本发明实施例的解像力测试方法中，首先以步长x来扫描图像，从而得到分辨率极限线LML的数值，但是，这只能够说明检测到的分辨率极限线LML在该步长的范围之内，还并无法准确地判定出分辨率极限线在哪一行，因此，可以从该分辨率极限线的-x到+x的范围内，进一步进行逐行扫描，从而确定准确的分辨率极限线LML的数值。这样，逐行扫描的计算量从原来的测试区域中整个楔形线图案的高度范围减小为两倍步长的范围，从而使得计算量大大减小。
另外，在根据本发明实施例的解像力测试方法中，为了精确地检测楔形线的开始行和结束行，当以步长x来扫描图像时，优选地也在楔形线的开始行和结束行的-x到+x的范围内进一步进行逐行扫描，从而精确地确定楔形线的开始行和结束行的数值。或者，可以首先逐行扫描图像，当检测到楔形线的开始行之后再以步长x来进行扫描。又或者，可以在以步长x检测到分辨率极限线之后，逐行扫描图像，以检测楔形线的结束行。
这里，本领域技术人员可以理解，当步长x的数值较大时，粗检步骤所需的计算量较小，但是细捡步骤所需的计算量较大，而当步长x的数值较小时，粗检步骤所需的计算量较大，但是细捡步骤所需的计算量较小，因此可以根据具体情况来灵活地设置粗检步骤的步长，仅需要满足上述关系即可。
根据本发明实施例的解像力测试方法简单易操作，并具有较高的运算效率， 算法精准，测试结果准确，能够较好的反应镜头的解像力，适于在生产线上批量使用，为镜头的解像力测试提供了一种简便易行、精准度较高的方法。
根据本发明的另一优选实施例，提供了一种解像力测试装置，与镜头模组连接以测试该镜头模组的解像力，其中所述解像力测试装置包括：
图像获取单元，用于通过该镜头模组获取至少一图像；
区域提取单元，用于从该图像获取单元所获取的图像中识别出至少一测试区域；
分辨率计算单元，用于计算该区域提取单元所提取的该测试区域中的分辨率值。
在上述解像力测试装置中，该区域提取单元具体用于，定位出该图像的四条侧边的中心位置O1，O2，O3，O4；将水平方向上两点的横坐标的差值定义为水平方向宽度，即W＝O3x-O1x，并垂直方向上两点的纵坐标的差值定义为垂直方向高度，即H＝O4y-O2y；基于测试区域中心与图像中心在水平和垂直方向上的距离占W/2和H/2的比例确定每个测试区域的中心坐标。
在上述解像力测试装置中，进一步包括去噪单元，用于对图像获取单元所获取的图像进行去噪。
优选地，该去噪单元仅对该区域提取单元所提取的测试区域进行去噪。
在上述解像力测试装置中，该测试区域中包括一楔形线图案，且该分辨率计算单元用于通过扫描该楔形线图案来计算分辨率值。
在上述解像力测试装置中，该区域提取单元具体包括：定位模块，用于自动定位出该图像中多个标记点的中心位置；提取模块，用于根据楔形线图案中心在连接所述多个标记点的多条线段上所占的比例不同，自动识别出楔形线图案。
在上述解像力测试装置中，进一步包括角度校正单元，用于计算出该图像与水平方向的夹角，并在该夹角不为零的情况下，作为该图像与镜头模组之间的偏转角度；且该分辨率计算单元用于沿着所述偏转角度扫描该楔形线图案以计算分辨率值。
在上述解像力测试装置中，进一步包括亮度调整单元，具体包括：亮度检测模块，用于检测该图像的整体亮度；亮度比较模块，用于将所检测的亮度与一标准亮度进行对比；和亮度增强模块，用于在所检测的亮度小于该标准亮度的情况下，对图像进行增强处理。
优选地，该亮度检测模块通过获取该图像的亮度Y分量来检测图像的整体亮度。
优选地，该亮度增强模块采用自适应图像增强技术对图像进行增强处理。
在上述解像力测试装置中，该分辨率计算单元对于由该去噪单元进行去噪，且由该亮度调整单元进行增强处理之后的图像，沿着由该角度校正单元计算的偏转角度扫描该楔形线图案以计算分辨率值。
在上述解像力测试装置中，该去噪单元、该角度校正单元和该亮度调整单元中的至少两个可以并行操作，或者该去噪单元、该角度校正单元和该亮度调整单元可以以任意顺序依此操作。
在上述解像力测试装置中，该分辨率计算单元从该楔形线图案较宽的一端开始，以垂直于该楔形线图案中的线条的方向扫描该楔形线图案，以通过楔形线起始行、楔形线分辨极限行和楔形线结束行来计算分辨率值。
具体地，该分辨率计算单元从该楔形线图案较宽的一端开始，以垂直于该楔形线图案中的线条的方向对该楔形线图案进行行扫描；在扫描的每一行中通过所扫描的像素点的灰度确定该行中所包含的楔形线的数目；并在当前行中所确定的楔形线数目小于前一行中所确定的楔形线数目的情况下，确定当前行为该分辨极限行。
优选地，该分辨率计算单元具体用于：以步长x扫描图像，以检测出初始楔形线起始行、初始楔形线分辨极限行和初始楔形线结束行；分别在初始楔形线起始行、初始楔形线分辨极限行和初始楔形线结束行的-x到+x的范围内逐行扫描，以检测出最终楔形线起始行、最终楔形线分辨极限行和最终楔形线结束行。
优选地，该分辨率计算单元具体用于：逐行扫描图像，以检测出最终楔形线起始行；以步长x扫描图像，以检测出初始楔形线分辨极限行；在初始楔形线分辨极限行的-x到+x的范围内逐行扫描，以检测出最终楔形线分辨极限行。
优选地，该分辨率计算单元具体用于：以步长x扫描图像，以检测出初始楔形线分辨极限行；在初始楔形线分辨极限行的-x到+x的范围内逐行扫描，以检测出最终楔形线分辨极限行；逐行扫描图像，以检测出最终楔形线结束行。
优选地，该步长x满足F(x)＝2x2-ax+a＜0，a为区域扫描范围。
当然，本领域技术人员可以理解，在根据本发明实施例的解像力测试装置中，也可以通过其他方法，例如计算机人工视觉算法来识别出楔形线图案的起始行、 分辨极限行和结束行，从而计算分辨率值。
并且，无论采用何种计算机人工视觉算法，均确定各个扫描的行中首次检测到线数减少的行为分辨极限行，并且，无论采用何种计算机人工视觉算法，均从楔形线图案的低频侧，即楔形线图案较宽的一段开始行扫描过程，从而与ISO 12233标准相一致。
这里，本领域技术人员可以理解，根据本发明实施例的解像力测试装置包括如图1所示的图像采集系统40和测试软件50，其中图像获取单元具体表现为图像采集系统40，而其它单元可以具体表现为测试软件50的形式。该测试软件50可以是机器代码、固件、嵌入代码和应用软件。另外，根据本发明实施例的解像力测试装置也可以实现为硬件，例如，硬件可以是电路、处理器、计算机、集成电路、集成电路内核、微型机电系统(MEMS)、无源器件或者其组合。
另外，测试软件50可以具体表现为计算机可读记录介质上的计算机可读代码。计算机可读记录介质是可以存储此后可以由计算机系统读取的数据的任何数据存储设备。计算机可读记录介质的实例包括只读存储器(ROM)、随机存取存储器(RAM)、CD-ROM、磁带、软盘、光数据存储装置和载波(比如通过因特网的数据传输)。计算机可读记录介质也可以经网络耦合的计算机系统分布，以使得以分布方式存储和执行计算机可读代码。
可以理解根据本发明的实施例的方法和装置可以由硬件、软件和/或其组合实现。软件可以存储在非易失性存储设备中，例如，可擦除或者可重写的只读存储器(ROM)、存储器，例如，随机存取存储器(RAM)、存储器芯片、存储器器件或者存储器集成电路(IC)、或者光学地或者磁性地可记录非瞬时机器可读的，例如，计算机可读的存储介质中，例如，致密盘(CD)、数字多用途盘(DVD)、磁盘或者磁带。根据本发明的实施例的方法和装置可以由包括控制器和存储器的计算机或者移动终端实现，且存储器可以是适于存储包括用于实现本发明的各种实施例的指令的一个或多个程序的非瞬时机器可读的，例如，计算机可读的存储介质的实例。
此外，根据本发明实施例的解像力测试装置可以电连接到图1所示的光源20，且亮度调整单元可以通过调整光源20的亮度来增强图像的亮度。
根据本发明的另一优选实施例，提供了一种解像力测试方法，用于测试镜头模组的解像力，其中所述解像力测试方法包括：
通过该镜头模组获取至少一图像；
从所获取的图像中识别出至少一测试区域；
计算所提取的该测试区域中的分辨率值。
在上述解像力测试方法中，在识别出至少一测试区域之后且在计算该测试区域中的分辨率值之前进一步包括：对所获取的图像进行去噪。
优选地，仅对所提取的测试区域进行去噪。
在上述解像力测试方法中，该测试区域中包括一楔形线图案，且计算所提取的该测试区域中的分辨率值具体为通过扫描该楔形线图案来计算所提取的该测试区域中的分辨率值。
在上述解像力测试方法中，从所获取的图像中识别出至少一测试区域具体包括：自动定位出该图像中多个标记点的中心位置；根据楔形线图案中心在连接所述多个标记点的多条线段上所占的比例不同，自动识别出楔形线图案。
在上述解像力测试方法中，在识别出至少一测试区域之后且在计算该测试区域中的分辨率值之前进一步包括：计算出该图像与水平方向的夹角，并在该夹角不为零的情况下，作为该图像与镜头模组之间的偏转角度；且计算所提取的该测试区域中的分辨率值具体为沿着所述偏转角度扫描该楔形线图案以计算分辨率值。
在上述解像力测试方法中，在识别出至少一测试区域之后且在计算该测试区域中的分辨率值之前包括：检测该图像的整体亮度；将所检测的亮度与一标准亮度进行对比；和在所检测的亮度小于该标准亮度的情况下，对图像进行增强处理。
优选地，通过获取该图像的亮度Y分量来检测图像的整体亮度。
优选地，采用自适应图像增强技术对图像进行增强处理。
在上述解像力测试装置中，该去噪步骤、该角度校正步骤和该亮度调整步骤中的至少两个可以并行进行，或者该去噪步骤、该角度校正步骤和该亮度调整步骤可以以任意顺序依此进行。
在上述解像力测试方法中，计算所提取的该测试区域中的分辨率值具体为：从该楔形线图案较宽的一段开始，以垂直于该楔形线图案中的线条的方向扫描该楔形线图案，以通过楔形线起始行、楔形线分辨极限行和楔形线结束行来计算分辨率值。
具体地，该分辨率计算步骤从该楔形线图案较宽的一端开始，以垂直于该楔 形线图案中的线条的方向对该楔形线图案进行行扫描；在扫描的每一行中通过所扫描的像素点的灰度确定该行中所包含的楔形线的数目；并在当前行中所确定的楔形线数目小于前一行中所确定的楔形线数目的情况下，确定当前行为该分辨极限行。
优选地，计算所提取的该测试区域中的分辨率值具体为：以步长x扫描图像，以检测出初始楔形线起始行、初始楔形线分辨极限行和初始楔形线结束行；分别在初始楔形线起始行、初始楔形线分辨极限行和初始楔形线结束行的-x到+x的范围内逐行扫描，以检测出最终楔形线起始行、最终楔形线分辨极限行和最终楔形线结束行。
优选地，计算所提取的该测试区域中的分辨率值具体为：逐行扫描图像，以检测出最终楔形线起始行；以步长x扫描图像，以检测出初始楔形线分辨极限行；在初始楔形线分辨极限行的-x到+x的范围内逐行扫描，以检测出最终楔形线分辨极限行。
优选地，计算所提取的该测试区域中的分辨率值具体为：以步长x扫描图像，以检测出初始楔形线分辨极限行；在初始楔形线分辨极限行的-x到+x的范围内逐行扫描，以检测出最终楔形线分辨极限行；逐行扫描图像，以检测出最终楔形线结束行。
优选地，该步长x满足F(x)＝2x2-ax+a＜0，a为区域扫描范围。
本发明的再一优选实施例提供了一种去噪方法，适于应用于解像力测试过程中，其中该去噪方法为局部去噪。
其中在去噪时，只对计算区域进行去噪。
本发明的又一优选实施例提供了一种图像亮度增强方法，其中该方法包括以下步骤：检测图像亮度；与一标准亮度进行对比；和在图像亮度低于该标准亮度的情况下，对图像亮度进行增强处理。
其中在检测到的图像亮度不低于标准亮度的情况下，不对图像亮度进行增强处理。
其中在上述方法中，采用的是自适应图像增强技术。
优选地，该标准亮度为预设亮度，预先置于一图像采集系统。
优选地，该标准亮度为适于测试的亮度。
本领域的技术人员应理解，上述描述及附图中所示的本发明的实施例只作为 举例而并不限制本发明。本发明的目的已经完整并有效地实现。本发明的功能及结构原理已在实施例中展示和说明，在没有背离所述原理下，本发明的实施方式可以有任何变形或修改。

Claims (27)

  1. 一种解像力测试方法，适于测试镜头的解像力，其特征在于，所述方法包括以下步骤：
    (A)获取至少一图像；
    (B)识别出至少一测试区域；
    (C)去噪；以及
    (D)计算分辨率值。
  2. 根据权利要求1所述的方法，其中在所述步骤(C)中，只对所述测试区域进行去噪。
  3. 根据权利要求1所述的方法，其中所述步骤(B)包括以下步骤：(B.1)自动定位出一标版上的多个标记点的中心位置；(B.2)计算出所述标版与水平方向的夹角，确定所述标版与一模组之间的偏转角度；以及(B.3)沿着所述偏转角度进行分辨率扫描。
  4. 根据权利要求2所述的方法，其中所述步骤(B)包括以下步骤：(B.1)自动定位出一标版上的多个标记点的中心位置；(B.2)计算出所述标版与水平方向的夹角，确定所述标版与一模组之间的偏转角度；以及(B.3)沿着所述偏转角度进行分辨率扫描。
  5. 根据权利要求1所述的方法，进一步包括一步骤(E)：检测图像亮度，并与一标准亮度进行对比，其中所述步骤(E)位于所述步骤(D)之前，并位于所述步骤(B)之后，当检测到的图像整体亮度低于所述标准亮度时，则对图像进行增强处理。
  6. 根据权利要求4所述的方法，进一步包括一步骤(E)：检测图像亮度，其中所述步骤(E)位于所述步骤(D)之前，并位于所述步骤(B)之后，当检测的图像整体亮度低于一标准亮度时，则对图像进行增强处理。
  7. 根据权利要求5所述的方法，其中在所述步骤(E)中，采用自适应图像增强技术，对图像进行增强处理。
  8. 根据权利要求1至7中任一所述的方法，其中所述步骤(D)包括以下步骤：(D.1)对图像进行分割，检测所述测试区域的楔形线的起始行和结束行；(D.2)粗检；(D.3)细检；以及(D.4)计算出线对数。
  9. 根据权利要求8所述的方法，其中在所述步骤(D.2)和所述步骤(D.3)中，所述粗检步骤适于快速定位到可分辨位置的附近，通过所述细检步骤进行逐行扫描，精确定位到最终的分辨率位置。
  10. 根据权利要求9所述的方法，其中在上述方法中，所述粗检步长需要满足如下公式：F(x)＝2x2-ax+a；其中a为区域扫描范围。
  11. 根据权利要求1所述的方法，其中所述测试区域中包括一楔形线图案，且计算所提取的该测试区域中的分辨率值具体为通过扫描所述楔形线图案来计算所提取的所述测试区域中的分辨率值。
  12. 根据权利要求11所述的方法，其中在上述解像力测试方法中，从所获取的图像中识别出至少一测试区域具体包括：自动定位出所述图像中多个标记点的中心位置；根据所述楔形线图案中心在连接所述多个标记点的多条线段上所占的比例不同，自动识别出所述楔形线图案。
  13. 根据权利要求12所述的方法，其中在识别出至少一所述测试区域之后且在计算该测试区域中的分辨率值之前进一步包括：计算出该图像与水平方向的夹角，并在所述夹角不为零的情况下，作为所述图像与镜头模组之间的偏转角度；且计算所提取的所述测试区域中的分辨率值具体为沿着所述偏转角度扫描所述楔形线图案以计算分辨率值。
  14. 根据权利要求13所述的方法，其中在识别出至少一所述测试区域之后且在计算所述测试区域中的分辨率值之前包括：检测所述图像的整体亮度；将所 检测的亮度与一标准亮度进行对比；和在所检测的亮度小于所述标准亮度的情况下，对图像进行增强处理。
  15. 根据权利要求11至14中任一所述的方法，其中通过获取该图像的亮度Y分量来检测图像的整体亮度。
  16. 根据权利要求11至14中任一所述的方法，其中计算所提取的所述测试区域中的分辨率值具体为：从所述楔形线图案较宽的一段开始，以垂直于所述楔形线图案中的线条的方向扫描所述楔形线图案，以通过楔形线起始行、楔形线分辨极限行和楔形线结束行来计算分辨率值。
  17. 根据权利要求16所述的方法，其中所述分辨率计算步骤从所述楔形线图案较宽的一端开始，以垂直于所述楔形线图案中的线条的方向对所述楔形线图案进行行扫描；在扫描的每一行中通过所扫描的像素点的灰度确定该行中所包含的楔形线的数目；并在当前行中所确定的楔形线数目小于前一行中所确定的楔形线数目的情况下，确定当前行为所述分辨极限行。
  18. 根据权利要求16所述的方法，其中计算所提取的所述测试区域中的分辨率值具体为：以步长x扫描图像，以检测出初始楔形线起始行、初始楔形线分辨极限行和初始楔形线结束行；分别在初始楔形线起始行、初始楔形线分辨极限行和初始楔形线结束行的-x到+x的范围内逐行扫描，以检测出最终楔形线起始行、最终楔形线分辨极限行和最终楔形线结束行。
  19. 根据权利要求16所述的方法，其中计算所提取的所述测试区域中的分辨率值具体为：逐行扫描图像，以检测出最终楔形线起始行；以步长x扫描图像，以检测出初始楔形线分辨极限行；在初始楔形线分辨极限行的-x到+x的范围内逐行扫描，以检测出最终楔形线分辨极限行。
  20. 根据权利要求16所述的方法，其中计算所提取的所述测试区域中的分辨率值具体为：以步长x扫描图像，以检测出初始楔形线分辨极限行；在初始楔 形线分辨极限行的-x到+x的范围内逐行扫描，以检测出最终楔形线分辨极限行；逐行扫描图像，以检测出最终楔形线结束行。
  21. 一种去噪方法，适于应用于解像力测试过程中，其特征在于，所述去噪方法为局部去噪。
  22. 根据权利要求21所述的去噪方法，其中去噪时，只对计算区域进行去噪。
  23. 一种图像亮度增强方法，其特征在于，所述方法包括以下步骤：(1)检测图像亮度；(2)与一标准亮度进行对比；以及(3)当图像亮度低于所述标准亮度时，对图像亮度进行增强处理。
  24. 根据权利要求23所述的方法，其中当检测到的图像亮度不低于于所述标准亮度时，则不需执行所述步骤(3)。
  25. 根据权利要求23所述的方法，其中在上述方法中，采用的是自适应图像增强技术。
  26. 根据权利要求23至25任一所述的方法，其中在上述方法中，所述标准亮度为预设亮度，预先置于一图像采集系统。
  27. 根据权利要求26所述的方法，其中在上述方法中，所述标准亮度为适于测试的亮度。
PCT/CN2016/093018 2015-08-04 2016-08-03 解像力测试方法和解像力测试装置 WO2017020829A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510470959.6 2015-08-04
CN201510470959.6A CN106441804B (zh) 2015-08-04 2015-08-04 解像力测试方法

Publications (1)

Publication Number Publication Date
WO2017020829A1 true WO2017020829A1 (zh) 2017-02-09

Family

ID=57942462

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/093018 WO2017020829A1 (zh) 2015-08-04 2016-08-03 解像力测试方法和解像力测试装置

Country Status (2)

Country Link
CN (1) CN106441804B (zh)
WO (1) WO2017020829A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108449476A (zh) * 2018-06-05 2018-08-24 昆山丘钛微电子科技有限公司 一种测试设备
CN109472767A (zh) * 2018-09-07 2019-03-15 浙江大丰实业股份有限公司 舞台灯具缺失状态分析系统
CN113566744A (zh) * 2021-07-29 2021-10-29 大连探索者科技有限公司 一种光电角度传感器分辨率的高精度测试方法

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108174189A (zh) * 2017-12-27 2018-06-15 清研新能源汽车工程中心（襄阳）有限公司 一种基于iso12233的解像力测试方法及系统
CN108827599A (zh) * 2018-05-21 2018-11-16 余姚舜宇智能光学技术有限公司 一种广角模组全视场解像力测试设备及方法
CN110896469B (zh) * 2018-09-13 2022-05-27 宁波舜宇光电信息有限公司 用于三摄的解像力测试方法及其应用
CN109308702A (zh) * 2018-09-14 2019-02-05 南京理工技术转移中心有限公司 一种目标实时识别定位方法
CN109191454A (zh) * 2018-09-14 2019-01-11 南京理工技术转移中心有限公司 一种电力监控系统的工作方法
CN109319619A (zh) * 2018-09-14 2019-02-12 南京理工技术转移中心有限公司 一种电梯智能监控系统的工作方法
CN109270853A (zh) * 2018-09-14 2019-01-25 南京理工技术转移中心有限公司 一种智能生态水资源管理系统的工作方法
CN109345541A (zh) * 2018-09-14 2019-02-15 南京理工技术转移中心有限公司 一种智能监控系统的工作方法
CN109064501A (zh) * 2018-09-14 2018-12-21 南京理工技术转移中心有限公司 一种污水处理监控系统的工作方法
CN110261069B (zh) * 2019-06-21 2022-01-18 舜宇光学(中山)有限公司 一种用于光学镜头的检测方法

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5818572A (en) * 1996-04-24 1998-10-06 Northrop Grumman Corporation Two-dimensional modulation transfer function measurement technique
CN201364235Y (zh) * 2008-12-30 2009-12-16 上海徕木电子股份有限公司 测试光电成像系统解像力的电视线图表板
CN102759440A (zh) * 2011-04-27 2012-10-31 华晶科技股份有限公司 解像力测试装置及其方法
CN103295191A (zh) * 2013-04-19 2013-09-11 北京航科威视光电信息技术有限公司 多尺度视觉自适应图像增强方法及评价方法
CN103903232A (zh) * 2014-04-10 2014-07-02 北京工业大学 在小波域利用进化规划进行图像去噪和增强的方法
CN104036472A (zh) * 2013-03-06 2014-09-10 北京三星通信技术研究有限公司 用于3d图像质量增强的方法和设备
CN104680485A (zh) * 2013-11-27 2015-06-03 展讯通信（上海）有限公司 一种基于多分辨率的图像去噪方法及装置
CN104751415A (zh) * 2013-12-31 2015-07-01 展讯通信（上海）有限公司 一种图像去噪和增强的方法、装置及图像处理系统

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN2601399Y (zh) * 2003-01-21 2004-01-28 明基电通股份有限公司 光学测量装置
CN101840576B (zh) * 2010-05-12 2011-12-21 浙江大学 一种可视化的测试数码相机各个成像区域分辨率的方法
CN101852675B (zh) * 2010-05-12 2011-12-07 浙江大学 一种测试数码相机各个成像区域分辨率的方法
JP5828649B2 (ja) * 2011-03-09 2015-12-09 キヤノン株式会社 画像処理装置、画像処理方法、及びコンピュータプログラム
CN103093419B (zh) * 2011-10-28 2016-03-02 浙江大华技术股份有限公司 一种检测图像清晰度的方法及装置
CN102663696B (zh) * 2012-03-31 2015-03-25 广东威创视讯科技股份有限公司 放大图像的去噪方法和系统
CN104156921B (zh) * 2014-08-08 2017-02-22 大连理工大学 一种低照度或亮度不均图像的自适应图像增强方法
CN104182998A (zh) * 2014-08-14 2014-12-03 深圳市云宙多媒体技术有限公司 一种自适应图像亮度渲染方法与装置

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5818572A (en) * 1996-04-24 1998-10-06 Northrop Grumman Corporation Two-dimensional modulation transfer function measurement technique
CN201364235Y (zh) * 2008-12-30 2009-12-16 上海徕木电子股份有限公司 测试光电成像系统解像力的电视线图表板
CN102759440A (zh) * 2011-04-27 2012-10-31 华晶科技股份有限公司 解像力测试装置及其方法
CN104036472A (zh) * 2013-03-06 2014-09-10 北京三星通信技术研究有限公司 用于3d图像质量增强的方法和设备
CN103295191A (zh) * 2013-04-19 2013-09-11 北京航科威视光电信息技术有限公司 多尺度视觉自适应图像增强方法及评价方法
CN104680485A (zh) * 2013-11-27 2015-06-03 展讯通信（上海）有限公司 一种基于多分辨率的图像去噪方法及装置
CN104751415A (zh) * 2013-12-31 2015-07-01 展讯通信（上海）有限公司 一种图像去噪和增强的方法、装置及图像处理系统
CN103903232A (zh) * 2014-04-10 2014-07-02 北京工业大学 在小波域利用进化规划进行图像去噪和增强的方法

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108449476A (zh) * 2018-06-05 2018-08-24 昆山丘钛微电子科技有限公司 一种测试设备
CN108449476B (zh) * 2018-06-05 2024-03-08 昆山丘钛微电子科技有限公司 一种测试设备
CN109472767A (zh) * 2018-09-07 2019-03-15 浙江大丰实业股份有限公司 舞台灯具缺失状态分析系统
CN109472767B (zh) * 2018-09-07 2022-02-08 浙江大丰实业股份有限公司 舞台灯具缺失状态分析系统
CN113566744A (zh) * 2021-07-29 2021-10-29 大连探索者科技有限公司 一种光电角度传感器分辨率的高精度测试方法

Also Published As

Publication number Publication date
CN106441804B (zh) 2019-08-30
CN106441804A (zh) 2017-02-22

Similar Documents

Publication Publication Date Title
WO2017020829A1 (zh) 解像力测试方法和解像力测试装置
CN109472271B (zh) 印刷电路板图像轮廓提取方法及装置
CN111179243A (zh) 一种基于计算机视觉的小尺寸芯片裂纹检测方法及系统
CN112686920A (zh) 一种圆形零件几何尺寸参数视觉测量方法及系统
CN106501272B (zh) 机器视觉焊锡定位检测系统
WO2017071406A1 (zh) 金针类元件的引脚检测方法和系统
CN116703909B (zh) 一种电源适配器生产质量智能检测方法
JP2018096908A (ja) 検査装置及び検査方法
WO2021195873A1 (zh) 识别sfr测试卡图像中感兴趣区域的方法及装置、介质
CN113176275B (zh) 一种用于显示面板复检的方法、装置及系统
CN116563298B (zh) 基于高斯拟合的十字线中心亚像素检测方法
CN113375555A (zh) 一种基于手机影像的电力线夹测量方法及系统
CN112669272A (zh) 一种aoi快速检测方法及快速检测系统
CN116908185A (zh) 物品的外观缺陷检测方法、装置、电子设备及存储介质
CN116337869A (zh) 一种基于机器视觉的双面线路板生产用检测系统
CN109448012A (zh) 一种图像边缘检测方法及装置
JP4821647B2 (ja) 電子部品の端子位置検出方法
KR20220154345A (ko) 가우시안 가중치 최소 자승법을 이용하는 영상 처리 장치 및 그것의 윤곽선 검출 방법
JP4247993B2 (ja) 画像検査装置、画像検査方法、制御プログラムおよび可読記憶媒体
TWI757590B (zh) 輝度均勻檢測系統及輝度均勻檢測方法
CN113012121A (zh) 裸片扫描结果的处理方法、装置、电子设备与存储介质
Zhu et al. Image quality evaluation method for surface crack detection based on standard test chart
JP4967132B2 (ja) 対象物表面の欠陥検査方法
CN116046791B (zh) 点胶缺陷检测方法及检测装置
CN110458840B (zh) 一种降低面板缺陷过检率的方法、系统及终端设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16832317

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16832317

Country of ref document: EP

Kind code of ref document: A1