WO2019148642A1 - 广角镜头的畸变矫正方法、装置及系统 - Google Patents

广角镜头的畸变矫正方法、装置及系统 Download PDF

Info

Publication number
WO2019148642A1
WO2019148642A1 PCT/CN2018/081908 CN2018081908W WO2019148642A1 WO 2019148642 A1 WO2019148642 A1 WO 2019148642A1 CN 2018081908 W CN2018081908 W CN 2018081908W WO 2019148642 A1 WO2019148642 A1 WO 2019148642A1
Authority
WO
WIPO (PCT)
Prior art keywords
view
angle
wide
field
image height
Prior art date
Application number
PCT/CN2018/081908
Other languages
English (en)
French (fr)
Inventor
王卓
刘绪明
曾吉勇
Original Assignee
江西联创电子有限公司
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 江西联创电子有限公司 filed Critical 江西联创电子有限公司
Priority to US16/195,284 priority Critical patent/US10748258B2/en
Publication of WO2019148642A1 publication Critical patent/WO2019148642A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/80—Geometric correction
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/70—SSIS architectures; Circuits associated therewith
    • H04N25/76—Addressed sensors, e.g. MOS or CMOS sensors
    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B13/00—Optical objectives specially designed for the purposes specified below
    • G02B13/04—Reversed telephoto objectives
    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B27/00—Optical systems or apparatus not provided for by any of the groups G02B1/00 - G02B26/00, G02B30/00
    • G02B27/0025—Optical systems or apparatus not provided for by any of the groups G02B1/00 - G02B26/00, G02B30/00 for optical correction, e.g. distorsion, aberration
    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B9/00—Optical objectives characterised both by the number of the components and their arrangements according to their sign, i.e. + or -
    • G02B9/64—Optical objectives characterised both by the number of the components and their arrangements according to their sign, i.e. + or - having more than six components
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/50—Constructional details
    • H04N23/54—Mounting of pick-up tubes, electronic image sensors, deviation or focusing coils
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/50—Constructional details
    • H04N23/55—Optical parts specially adapted for electronic image sensors; Mounting thereof
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/80—Camera processing pipelines; Components thereof
    • H04N23/81—Camera processing pipelines; Components thereof for suppressing or minimising disturbance in the image signal generation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/60—Noise processing, e.g. detecting, correcting, reducing or removing noise
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/60—Noise processing, e.g. detecting, correcting, reducing or removing noise
    • H04N25/61—Noise processing, e.g. detecting, correcting, reducing or removing noise the noise originating only from the lens unit, e.g. flare, shading, vignetting or "cos4"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30248—Vehicle exterior or interior
    • G06T2207/30252—Vehicle exterior; Vicinity of vehicle

Definitions

  • the present invention relates to the field of image processing technologies, and in particular, to a distortion correction method, device and system for a wide-angle lens.
  • the ADAS Advanced Driver Assistance System
  • the technical treatment enables the driver to detect the possible dangers in the fastest time and effectively increase the comfort and safety of the driving. Therefore, ADAS plays a very important role in the process of achieving automatic driving.
  • a typical ADAS system generally has three cameras at the front end of the vehicle, and the field of view angles are approximately 42 degrees, 60 degrees, and 120 degrees. If binocular vision is required, the number of cameras needs to be doubled; each camera is equipped with one 2MP (Mega-Pixels, megapixels), a 2 megapixel CMOS sensor.
  • 2MP Mega-Pixels, megapixels
  • the settings in the related art have great redundancy.
  • the field of view of the 42-degree camera itself is covered by the 60-degree camera and the 120-degree camera field of view, so that multiple cameras not only increase the complexity of the system connection but also the various cameras.
  • the complexity of communication and synchronization, and the cost of the entire system increases, and the need to ensure a good overlap of the field of view between the cameras, resulting in a fixed angular relationship between each other, while the overall
  • the car itself must also guarantee the angle, so it is necessary to regularly correct between the camera and between the camera and the car body.
  • the three camera corrections themselves are relatively complicated and require a lot of work to be solved.
  • the present invention aims to solve at least one of the technical problems in the related art to some extent.
  • an object of the present invention is to provide a distortion correction method for a wide-angle lens, which can achieve a central field of view sampling rate of a wide-angle lens higher than an edge, so that one camera system can be used instead of three camera systems.
  • Another object of the present invention is to provide a distortion correcting device for a wide-angle lens.
  • an embodiment of the present invention provides a distortion correction method for a wide-angle lens, comprising the steps of: obtaining n half-field angles ⁇ 1 to ⁇ n of a wide-angle lens, wherein ⁇ n is ⁇ (n-1) ) adjacent half angle of view, and field of view of obtaining a ratio ⁇ 1 to ⁇ (n-1) according to the n adjacent to the half angle ⁇ 1 to [theta] n, n being a positive integer; in accordance with the n half field angles ⁇ 1 to ⁇ n obtain image heights r 1 to r n corresponding to the n half field angles; according to the adjacent field ratios ⁇ 1 to ⁇ (n-1) and The image heights r 1 to r n corresponding to the n half angles of view obtain the image height relationship of the adjacent half field of view, and the image height corresponding to the maximum angle of view and the image height of the adjacent half field of view The image height of each of the half angles of
  • the distortion correction method of the wide-angle lens according to the embodiment of the present invention can obtain the image height relationship of the adjacent half angle of view according to the image height corresponding to the adjacent field of view ratio and the plurality of half angles of view, and according to the image corresponding to the maximum angle of view.
  • the image height relationship between the high and adjacent half angles of view is obtained by recursively calculating the image height of each half angle of view, thereby correcting the distortion of the wide field lens in the central field of view according to the image height of each half angle of view, realizing a wide angle lens
  • the sampling rate of the central field of view is higher than the edge, which improves the accuracy and reliability of the correction, improves the correction efficiency, effectively reduces the cost, and is easy to implement.
  • the distortion correcting method of the wide-angle lens according to the above embodiment of the present invention may further have the following additional technical features:
  • the obtaining the n half field angles ⁇ 1 to ⁇ n of the wide-angle lens further includes: acquiring a maximum angle of view of the wide-angle lens; and determining a maximum field of view according to the wide-angle lens The angle obtains the half angle of view ⁇ 1 and obtains a half angle of view ⁇ n from the adjacent previous half angle of view angle ⁇ (n-1) .
  • the image heights r 1 to r n corresponding to the adjacent field of view ratios ⁇ 1 to ⁇ (n-1) and the n half field angles are obtained.
  • the image height relationship of the adjacent half angles of view further includes: sampling the image heights r 1 to r n corresponding to the n half field angles by a field angle relationship, wherein the sampling process uses interpolation Method; obtaining an image height relation of an adjacent half angle of view according to the angle of view relationship.
  • the image height relationship of the adjacent half angle of view is:
  • the image height of each of the half angles of view is obtained by recursively calculating according to an image height corresponding to a maximum angle of view and an image height relationship of the adjacent half field of view:
  • r (n-1) is the image height corresponding to the half angle of view ⁇ (n-1)
  • r n is the image height corresponding to the half angle of view ⁇ n
  • ⁇ 1 is the ratio of ⁇ 1 and ⁇ 2
  • ⁇ 2 is the ratio of ⁇ 2 and ⁇ 3
  • ⁇ (n-1) is the ratio of ⁇ (n-1) and ⁇ n .
  • the maximum angle of view is greater than or equal to 70°.
  • the maximum angle of view is greater than or equal to 100 degrees and less than or equal to 190 degrees.
  • another embodiment of the present invention provides a distortion correcting apparatus for a wide-angle lens, comprising: an acquisition module for acquiring n half-field angles ⁇ 1 to ⁇ n of a wide-angle lens, wherein ⁇ n is ⁇ An adjacent half angle of view of (n-1) , and acquiring the adjacent field ratios ⁇ 1 to ⁇ (n-1) according to the n half angles of view ⁇ 1 to ⁇ n , where n is a positive integer ; obtaining module, configured to obtain the n half angle of view corresponding to the image height r 1 to r n according to the n half field angle ⁇ 1 to [theta] n; calculating module, adjacent to the view according The image ratios ⁇ 1 to ⁇ (n-1) and the image heights r 1 to r n corresponding to the n half angles of view obtain image height relations of adjacent half angles of view, and images corresponding to the maximum angle of view Height and the image height relationship of the adjacent half angle
  • the distortion correcting device for the wide-angle lens of the embodiment of the present invention can obtain the image height relationship of the adjacent half angle of view according to the image height corresponding to the adjacent field of view ratio and the plurality of half angles of view, and according to the image corresponding to the maximum angle of view
  • the image height relationship between the high and adjacent half angles of view is obtained by recursively calculating the image height of each half angle of view, thereby correcting the distortion of the wide field lens in the central field of view according to the image height of each half angle of view, realizing a wide angle lens
  • the sampling rate of the central field of view is higher than the edge, which improves the accuracy and reliability of the correction, improves the correction efficiency, effectively reduces the cost, and is easy to implement.
  • the distortion correcting device of the wide-angle lens according to the above embodiment of the present invention may further have the following additional technical features:
  • the acquiring module further includes: an acquiring unit, configured to acquire a maximum angle of view of the wide-angle lens; and a selecting unit, configured to obtain, according to a maximum angle of view of the wide-angle lens the half angle of view ⁇ 1, and (n-1) to give the half field angle ⁇ n in accordance with an adjacent front half angle ⁇ .
  • the calculating module further includes: a processing unit, configured to perform image heights r 1 to r n corresponding to the n half field of view angles by a viewing angle relationship And a sampling process, wherein the sampling process adopts an interpolation method; and an acquiring unit is configured to obtain an image height relation of an adjacent half angle of view according to the angle of view relationship.
  • the image height relationship of the adjacent half angle of view is:
  • the image height of each of the half angles of view is obtained by recursively calculating according to an image height corresponding to a maximum angle of view and an image height relationship of the adjacent half field of view:
  • r (n-1) is the image height corresponding to the half angle of view ⁇ (n-1)
  • r n is the image height corresponding to the half angle of view ⁇ n
  • ⁇ 1 is the ratio of ⁇ 1 and ⁇ 2
  • ⁇ 2 is the ratio of ⁇ 2 and ⁇ 3
  • ⁇ (n-1) is the ratio of ⁇ (n-1) and ⁇ n .
  • the maximum angle of view is greater than or equal to 70°.
  • the maximum angle of view is greater than or equal to 100 degrees and less than or equal to 190 degrees.
  • a further aspect of the present invention provides a distortion correction system for a wide-angle lens, which includes the wide-angle lens and the pixel sensor corrected by the distortion correcting device of the wide-angle lens described above.
  • the distortion correction system of the wide-angle lens can obtain the image height relationship of the adjacent half angle of view according to the image height corresponding to the adjacent field of view ratio and the plurality of half angles of view, and according to the image corresponding to the maximum angle of view.
  • the image height relationship between the high and adjacent half angles of view is obtained by recursively calculating the image height of each half angle of view, thereby correcting the distortion of the wide field lens in the central field of view according to the image height of each half angle of view, realizing a wide angle lens
  • the sampling rate of the central field of view is higher than the edge, which improves the accuracy and reliability of the correction, improves the correction efficiency, effectively reduces the cost, and is easy to implement.
  • the wide-angle lens includes a first lens, a second lens, a third lens, a diaphragm, a fourth lens, a fifth lens, and a sixth lens, in order from the object side to the image side.
  • the seventh lens and the filter are the first lens, a second lens, a third lens, a diaphragm, a fourth lens, a fifth lens, and a sixth lens, in order from the object side to the image side.
  • the first lens, the second lens and the seventh lens each adopt an aspherical lens.
  • FIG. 1 is a flow chart of a distortion correction method of a wide-angle lens according to an embodiment of the present invention
  • FIG. 2 is a schematic diagram showing the image height relationship of a wide-angle lens at an adjacent field of view according to an embodiment of the present invention
  • FIG. 3 is a schematic structural view of a distortion correcting device for a wide-angle lens according to an embodiment of the present invention
  • FIG. 4 is a schematic structural view of a wide-angle lens using recursive distortion correction according to an embodiment of the present invention
  • FIG. 5 is a schematic diagram of a grid pattern of a wide-angle lens distortion correction system before downsampling according to an embodiment of the present invention
  • Figure 6 is a schematic diagram of a grid pattern of a wide-angle lens distortion correction system after downsampling, in accordance with an embodiment of the present invention.
  • the sensors on the market and the ADAS system are equipped with 2MP pixels, which are about 1/3" in size, and the next generation of sensors will be 8MP. Based on this, can you design a camera with a high pixel? The sensor camera achieves the effect of the existing three sensor cameras? If the simple superposition according to the pixels, the number of pixels of the latest 8MP sensor has exceeded the sum of the three 2MP sensors. However, the 8MP sensor can satisfy the maximum angle of view 120. The pixel requirement of °, but not necessarily the 2MP pixel requirement of the central field of view angle of 60° and 42°.
  • a typical imaging optical system has a problem that the sampling rate is low near the central field of view and the edge sampling rate is high, and the illuminance of the edge also decreases due to an increase in the angle of view.
  • the picture of 8MP at 120 degrees, after intercepting the center corresponding to the 60 degree area is 2MP, corresponding to the interception center
  • the 42-degree field of view is 1MP
  • the most important problem of the ideal distortion correction method and the f-theta distortion correction method in the related art is the excessively high sampling rate.
  • the edge sampling rate is higher than the central field of view, that is, the edge is too strong;
  • the edge and center sampling rate are the same, that is, uniform force.
  • the invention is based on the above problems, and proposes a distortion correction method and device for a wide-angle lens.
  • FIG. 1 is a flow chart of a distortion correcting method of a wide-angle lens according to an embodiment of the present invention.
  • the distortion correction method of the wide-angle lens includes the following steps:
  • step S101 n half-field angles ⁇ 1 to ⁇ n of the wide-angle lens are acquired, wherein ⁇ n is an adjacent half-field angle of ⁇ (n-1) , and according to n half-field angles ⁇ 1 to ⁇ n acquires adjacent field ratios ⁇ 1 to ⁇ (n-1) , where n is a positive integer.
  • acquiring the n half field angles ⁇ 1 to ⁇ n of the wide-angle lens further includes: acquiring a maximum angle of view of the wide-angle lens; and obtaining a half field angle according to the maximum angle of view of the wide-angle lens ⁇ 1, and in accordance with an adjacent front half angle ⁇ (n-1) to give the half field angle ⁇ n.
  • the maximum field of view is greater than or equal to 70°.
  • the maximum angle of view is greater than or equal to 100 degrees and less than or equal to 190 degrees.
  • the wide-angle lens also known as a short focal length lens, is a lens with a field of view angle greater than 60°; for a zoom lens on a camera, a focal length of less than 25 mm, the use of a wide-angle lens facilitates the display of a wide range of scenes at close range, and
  • the wide-angle lens is suitable for displaying the main body of the picture and the environment in which it is located.
  • the embodiment of the present invention can first obtain the maximum HFOV (Horizontal Field Of View) of the wide-angle lens, and sequentially select the half-field angles adjacent to the n wide-angle lenses, such as sequentially selecting the half-view adjacent to the wide-angle lens.
  • the maximum angle of view of the wide angle lens may be greater than or equal to 70°, and the maximum angle of view of the preferred wide-angle lens is greater than or equal to 100° and less than or equal to 190°.
  • step S102 the n according to the half field angle ⁇ 1 to obtain [theta] n n half angle of view corresponding to the image height r 1 to r n.
  • step S103 the image height relationship of the adjacent half field of view angle is obtained according to the adjacent image field ratios ⁇ 1 to ⁇ (n-1) and the image heights r 1 to r n corresponding to the n half field angles, and according to The image height relationship corresponding to the image height corresponding to the maximum angle of view and the adjacent half angle of view is obtained by recursively calculating the image height of each half angle of view.
  • adjacent half angles of view are obtained according to image heights r 1 to r n corresponding to adjacent field ratios ⁇ 1 to ⁇ (n-1) and n half angles of view
  • the image height relationship further includes: sampling the image heights r 1 to r n corresponding to the n half field angles by a field angle relationship, wherein the sampling process adopts an interpolation method; and the phase is obtained according to the angle of view field relationship.
  • the image height relationship of the adjacent half field of view is obtained according to image heights r 1 to r n corresponding to adjacent field ratios ⁇ 1 to ⁇ (n-1) and n half angles of view
  • the image height relationship further includes: sampling the image heights r 1 to r n corresponding to the n half field angles by a field angle relationship, wherein the sampling process adopts an interpolation method; and the phase is obtained according to the angle of view field relationship.
  • the image height relationship of adjacent half angles of view is:
  • the image height of each half angle of view is obtained by recursively calculating according to the image height corresponding to the maximum angle of view and the image height relationship of the adjacent half angle of view:
  • r (n-1) is the image height corresponding to the half angle of view ⁇ (n-1)
  • r n is the image height corresponding to the half angle of view ⁇ n
  • ⁇ 1 is the ratio of ⁇ 1 and ⁇ 2
  • ⁇ 2 is the ratio of ⁇ 2 and ⁇ 3
  • ⁇ (n-1) is the ratio of ⁇ (n-1) and ⁇ n .
  • the embodiment of the present invention can define the adjacent field of view ratio as follows:
  • ⁇ 1 ⁇ 1 / ⁇ 2
  • ⁇ 2 ⁇ 2 / ⁇ 3
  • ⁇ (n-1) ⁇ (n-1) / ⁇ n ,
  • the image heights of the n half field of view angles are sampled according to the ratio of the adjacent half angles of view to obtain a relational expression, wherein the image heights of the n half field of view angles of the wide-angle lens are sampled by interpolation.
  • a relational expression wherein the image heights of the n half field of view angles of the wide-angle lens are sampled by interpolation.
  • r 1 is the image height corresponding to the half angle of view ⁇ 1
  • r 2 is the image height corresponding to the half angle of view ⁇ 2
  • r 3 is the image height corresponding to the half angle of view ⁇ 3
  • r (n-1) For the image height corresponding to the half angle of view ⁇ (n-1)
  • r n is the image height corresponding to the half angle of view ⁇ n
  • ⁇ 1 is the ratio of ⁇ 1 and ⁇ 2
  • ⁇ 2 is ⁇ 2 and ⁇ 3
  • the ratio, ⁇ (n-1) is the ratio of ⁇ (n-1) and ⁇ n
  • n is an integer greater than or equal to 3.
  • the meaning of the relation (3) is the center of a small field of high resolution corresponding to image height ⁇ n r n
  • downsampling may be performed when adjacent large field of view ⁇ (n-1),
  • the sampling magnification is the field of view ratio ⁇ (n-1)
  • the newly added large field of view image height (r (n-1) - r n ) part does not need to be resampled under ideal conditions, ie no waste of existing
  • the pixel resolution the result is that the large field of view ⁇ (n-1) itself achieves a pixel resolution that is completely consistent with the adjacent small field of view ⁇ n , that is, the large field of view image height after downsampling is equal to r n .
  • the embodiment of the present invention can calculate the image height relationship by recursion:
  • step S104 the distortion of the wide-angle lens in the central field of view is sequentially corrected in accordance with the image height of each half-angle of view.
  • the embodiment of the present invention can sequentially correct the distortion of the wide-angle lens in the central field of view according to the image height of each half-angle of view.
  • the embodiment of the present invention adopts a recursive distortion corrected wide-angle lens system, and is preferably an optical system composed of a 8MP CMOS (Complementary Metal Oxide Semiconductor) sensor and a wide-angle lens having a maximum viewing angle of 120 degrees. It can realize the system of effectively replacing three lenses (the angle of view is 120 degrees, 60 degrees and 42 degrees respectively) and three 2MP sensors, that is, the field of view of the entire wide-angle lens is 120 degrees, and the corresponding resolution is 8MP.
  • the small field of view of the center of 60 degrees and 42 degrees can also achieve the resolution of the center 2MP; the specific steps are as follows:
  • the maximum angle of view of the wide angle lens of the wide angle lens is half of ⁇ 1 of 60 degrees, and the image height corresponding to the half angle of view is r 1 ; the half angle ⁇ 2 of the adjacent medium angle of view is 60 degrees. Degree, the image height corresponding to the half angle of view is r 2 ; the half angle ⁇ 3 of the central small field angle of 42 degrees is 21 degrees, and the image height corresponding to the half angle of view is r 3 ; thus the adjacent field of view ratio is obtained:
  • the image heights corresponding to the half angles of view ⁇ 2 and ⁇ 3 are respectively sampled and processed by the following relation:
  • the above-mentioned wide-angle lens is subjected to distortion correction design in accordance with the obtained image height of each half angle of view.
  • the beneficial effects brought by the recursive distortion correction method provided by the embodiment of the present invention and the common distortion correction method in the related art are shown in Table 1. Since this distortion requirement is based on nonlinear sampling of the system, it is not based on the human eye's own preferences, but is completely directed to machine vision, and the specific implementation of such sampling adopted by the embodiment of the present invention is interpolation.
  • the interpolation point itself is a non-linear distribution, so the interpolation position can be calculated in advance to increase the speed.
  • Table 1 is an image height comparison table of the recursive distortion control and the distortion control in the related art according to an embodiment of the present invention.
  • CMOS sensor with a pixel number of 8MP and a wide-angle lens with a maximum viewing angle of 120 degrees are used. If the image height relationship is converted into a pixel relationship, such as the AR0820 of the ON Semiconductor 8MP, the pixel is 3840*2160, then half. The pixel of the field of view is 1920*1080, and the correspondence between the recursive distortion control of the embodiment of the present invention and the distortion control pixel in the related art can be obtained, as shown in Table 2.
  • Table 2 is a horizontal pixel comparison table of the recursive distortion control and the distortion control in the related art according to an embodiment of the present invention.
  • the result of the recursive distortion correction control adopted by the embodiment of the present invention is that the central portion of the wide-angle lens can achieve a resolution of 2MP after interception.
  • the corresponding number of pixels is 985, which is substantially the same as the 964 (1928/2) pixel of AR0231 of ON Semiconductor 2MP. That is to say, the amount of pixel information contained in this part is the same as that of designing a lens separately and equipped with AR0231.
  • the effective pixel resolution after downsampling of the 30 degree field of view is also 985.
  • the wide-angle lens using nonlinear recursive distortion correction has a resolution of 8MP corresponding to a field angle of 120 degrees, a small field of view of 42 degrees of the center, and a resolution of the center 2MP;
  • the field is downsampled, and combined with the adjacent 60-degree field of view information, the resolution of 2MP can also be achieved; further downsampling the image of the 60-degree field of view, combined with information between 60 degrees and 120 degrees, can achieve full field of view. 2MP resolution.
  • embodiments of the present invention may replace multiple low pixel sensors and multiple lens systems with a higher pixel sensor and a recursively corrected wide angle lens.
  • a denoising algorithm can be added during the downsampling to solve the problem of increased dark noise due to pixel size reduction. For example, the binning of the central part of the pixel. For example, if it is twice the downsampling, you can simply make a 2 ⁇ 2 matrix kernel convolution on the center part and then sample it.
  • a typical application scenario is when the daytime light is sufficient, the central part such as the 42 degree field of view has sufficient resolution, the signal strength is sufficient, the ADAS system can look far; the night light is insufficient, due to the illumination limit of the street light, 42 degrees Although the camera looks far, the obtained photos are very noisy and the illumination is not enough.
  • the 60-degree field of view can be downsampled to remove some noise.
  • the resolution of the central portion within 42 degrees is sacrificed, the low light sensitivity of the system is improved. Since the driving speed at night can be lower, it is acceptable to have a lower resolution of the object to be discerned. Further, if the illuminance is lower and the visibility is worse, the noise and downsampling can be simultaneously performed on the center of 60 degrees and 42 degrees based on the case of the 120 degree field of view.
  • embodiments of the present invention can use an 8MP CMOS chip, which now produces three 2MP images corresponding to a 42 degree field of view, a 60 degree field of view, and a 120 degree field of view, respectively.
  • the ADAS system can be used with both the 42-degree field of view and the 120-degree field of view.
  • the 42-degree field of view ensures that the ADAS system can be viewed very far.
  • the 8MP sensor chip has a pixel size of 2.1um, which is equivalent to 2 ⁇ 2 binning for the sensor, and the effective pixel size is 4.2um. This size and low light performance is very good AR0220 chip. It is consistent.
  • the wide-angle lens distortion correction method of the embodiment of the invention has the following beneficial effects: (1)
  • the embodiment of the invention effectively corrects the distortion of the wide-angle lens, and achieves a high sampling rate of the central field of view.
  • the resolution of the central field of view of the wide-angle lens can be quantitatively adjusted as required by the system;
  • the contrast of the lens system can be even greater than 1, This makes it possible to increase the CRA (Chief Ray Angle) during optical design, thereby reducing the overall length of the system and making the entire system more compact.
  • CRA Choef Ray Angle
  • the distortion at 1280 pixels does not need to be strictly controlled.
  • the main reason is that it is close to the central field of view, that is, the 21 degree half field of view and the 30 degree half field of view are very close, and the effective control of the chief ray may have certain challenges.
  • the direct result is that when the 30-degree field of view is sampled, the size of the resulting picture exceeds 2MP (such as 3MP), which is easily further processed by subsequent digital processing. Correction;
  • 2MP such as 3MP
  • Correction Another compromise that can be considered is strict control of the 30-degree half field of view, while the 21-degree half field of view itself is more difficult to do distortion correction.
  • the consequence is that the picture cut at the 21-degree half-view site itself does not have 2MP.
  • the choice of specific solutions needs to be determined by the person skilled in the art according to the actual situation.
  • the distortion at 60 degrees is a key to realize the algorithm, and the distortion needs effective control here.
  • the image height relationship of the adjacent half field of view angle can be obtained according to the image height corresponding to the adjacent field of view ratio and the plurality of half angles of view, and corresponding to the maximum field of view angle.
  • the image height relationship between the image height and the adjacent half angle of view is obtained by recursively calculating the image height of each half angle of view, thereby correcting the distortion of the wide field lens in the central field of view according to the image height of each half angle of view.
  • the sampling rate of the central field of view of the wide-angle lens is higher than that of the edge, so that the resolution of the central field of view of the wide-angle lens can be quantitatively adjusted according to the system requirements, and the overall length of the system can be reduced by increasing the incident angle of the light to make the whole system more compact;
  • High-pixel cameras effectively replace multiple low-pixel cameras, saving the cost of the ADAS system while simplifying system correction and system calculation requirements, improving the accuracy and reliability of correction, improving correction efficiency, reducing costs, and easily implementing .
  • FIG. 3 is a schematic structural view of a distortion correcting device for a wide-angle lens according to an embodiment of the present invention.
  • the distortion correcting device 10 of the wide-angle lens includes an acquisition module 100, an acquisition module 200, a calculation module 300, and a correction module 400.
  • the acquisition module 100 is configured to acquire n half-field angles ⁇ 1 to ⁇ n of the wide-angle lens, where ⁇ n is an adjacent half-field angle of ⁇ (n-1) , and according to n half-field angles ⁇ 1 to ⁇ n acquire adjacent field ratios ⁇ 1 to ⁇ (n-1) , where n is a positive integer.
  • N obtaining module 200 to obtain a half angle of view corresponding to the image height r 1 to r n n The half field angle ⁇ 1 to ⁇ n.
  • the calculation module 300 is configured to obtain an image height relationship of adjacent half angles of view according to image heights r 1 to r n corresponding to adjacent field ratios ⁇ 1 to ⁇ (n-1) and n half angles of view, and according to The image height relationship corresponding to the image height corresponding to the maximum angle of view and the adjacent half angle of view is obtained by recursively calculating the image height of each half angle of view.
  • the correction module 400 is configured to sequentially correct the distortion of the wide-angle lens in the central field of view according to the image height of each half-angle of view.
  • the device 10 of the embodiment of the present invention can correct the distortion of the wide-angle lens in the central field of view according to the image height relationship of the adjacent half-angle of view, so that the sampling rate of the central field of view is higher than the edge, and the cost is reduced, which is simple and easy to implement.
  • the acquisition module 100 further includes: an acquisition unit and a selection unit.
  • the acquisition unit is used to collect the maximum angle of view of the wide-angle lens.
  • the selection unit is configured to obtain a half angle of view ⁇ 1 according to the maximum angle of view of the wide-angle lens, and obtain a half angle of view ⁇ n according to the adjacent previous half angle of view angle ⁇ (n-1) .
  • the calculation module 300 further includes: a processing unit and a calculation unit.
  • the processing unit is configured to perform sampling processing on the image heights r 1 to r n corresponding to the n half field of view angles by using a field of view relationship, wherein the sampling process uses an interpolation method.
  • the obtaining unit is configured to obtain an image height relation of adjacent half angles of view according to a viewing angle relationship.
  • the image height relationship of adjacent half angles of view is:
  • the image height of each half angle of view is obtained by recursively calculating according to the image height corresponding to the maximum angle of view and the image height relationship of the adjacent half angle of view:
  • r (n-1) is the image height corresponding to the half angle of view ⁇ (n-1)
  • r n is the image height corresponding to the half angle of view ⁇ n
  • ⁇ 1 is the ratio of ⁇ 1 and ⁇ 2
  • ⁇ 2 is the ratio of ⁇ 2 and ⁇ 3
  • ⁇ (n-1) is the ratio of ⁇ (n-1) and ⁇ n .
  • the maximum field of view is greater than or equal to 70°.
  • the maximum angle of view is greater than or equal to 100 degrees and less than or equal to 190 degrees.
  • the image height relationship of the adjacent half field of view angle can be obtained according to the image height corresponding to the adjacent field of view ratio and the plurality of half angles of view, and corresponding to the maximum field of view angle
  • the image height relationship between the image height and the adjacent half angle of view is obtained by recursively calculating the image height of each half angle of view, thereby correcting the distortion of the wide field lens in the central field of view according to the image height of each half angle of view.
  • the sampling rate of the central field of view of the wide-angle lens is higher than that of the edge, so that the resolution of the central field of view of the wide-angle lens can be quantitatively adjusted according to the system requirements, and the overall length of the system can be reduced by increasing the incident angle of the light to make the whole system more compact.
  • the embodiment of the present invention further provides a distortion correction system for a wide-angle lens, which includes the wide-angle lens and the pixel sensor corrected by the distortion correction device of the wide-angle lens described above.
  • the wide-angle lens corrected by the distortion correcting device includes a first lens, a second lens, a third lens, a diaphragm, a fourth lens, and a fifth from the object side to the image side in order.
  • the first lens, the second lens and the seventh lens each employ an aspherical lens.
  • the aspherical surface shapes of the first lens, the second lens, and the seventh lens all satisfy the following equation:
  • z is the curvature corresponding to the radius
  • h is the radial coordinate
  • c is the curvature of the vertices of the surface
  • K is the conic quadratic coefficient
  • B, C, D, and E represent the fourth, sixth, eighth, and ten, respectively.
  • the recursive distortion corrected wide-angle lens (the maximum angle of view is 120 degrees) used in the embodiment of the present invention includes, from the object side to the imaging surface, a first lens L1 and a second lens L2.
  • the third lens L3, the aperture S1, the fourth lens L4, the fifth lens L5, the sixth lens L6, the seventh lens L7, and the filter S2, and the first lens L1, the second lens L2, and the seventh lens L7 are Aspherical lens.
  • the design parameters of each lens in the wide-angle lens of the embodiment of the present invention are shown in Table 3.
  • the aspherical parameters of each lens are shown in Table 4.
  • the small x is a uniform sampling of the main ray to the field of view.
  • the deformation occurs as a whole due to the distortion, and the result is that the central portion is not too distorted, a small x can correspond to a sample of about 9 pixels, so the resolution is high, and a small x on the side corresponds to only one pixel of the sample, the resolution is low, and the small black square in the figure represents one pixel.
  • these small x will be evenly distributed, and the reconstructed image is shown in Figure 6.
  • Table 5 is the corresponding image height and pixel parameters of the wide-angle lens provided by the embodiment at different half angles of view.
  • the theoretical values of the wide-angle lens of the embodiment of the present invention and the recursive distortion correction method of the present invention are basically the same. It is to be noted that the recursive distortion correction method of the embodiment of the invention can effectively correct the distortion of the wide-angle lens in the central field of view, and can realize the sampling rate of the central field of view is higher than the sampling rate of the edge.
  • the refraction distortion corrected wide-angle lens by adopting a refraction distortion corrected wide-angle lens, and assuming a system of a 160-degree wide-angle lens and a 12.4 MP CMOS sensor, if the resolution of the center 42-degree field of view is still maintained For 2MP, then according to Table 6, if the lateral field of view is increased to 160 degrees, the sensor needs to have a pixel of 4802*2702, which is 12.4MP. In addition, if the horizontal field of view is kept constant at 120 degrees, the number of pixels of the central 42-degree field of view is also increased to 3 MP after the sensor is increased to 12 MP.
  • the development of the next generation wide-angle lens products can have two directions, one is to improve the field of view coverage and maintain the central resolution; the other is to maintain the field of view coverage and improve the central resolution. It should be noted that the embodiment of the present invention can determine a more suitable solution according to the development and algorithm of the ADAS.
  • the image height relationship of the adjacent half field of view angle can be obtained according to the image height corresponding to the adjacent field of view ratio and the plurality of half angles of view, and corresponding to the maximum field of view angle.
  • the image height relationship between the image height and the adjacent half angle of view is obtained by recursively calculating the image height of each half angle of view, thereby correcting the distortion of the wide field lens in the central field of view according to the image height of each half angle of view.
  • the sampling rate of the central field of view of the wide-angle lens is higher than that of the edge, so that the resolution of the central field of view of the wide-angle lens can be quantitatively adjusted according to the system requirements, and the overall length of the system can be reduced by increasing the incident angle of the light to make the whole system more compact;
  • High-pixel cameras effectively replace multiple low-pixel cameras, saving the cost of the ADAS system while simplifying system correction and system calculation requirements, improving the accuracy and reliability of correction, improving correction efficiency, reducing costs, and easily implementing .
  • first and second are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated.
  • features defining “first” or “second” may include at least one of the features, either explicitly or implicitly.
  • the meaning of "a plurality” is at least two, such as two, three, etc., unless specifically defined otherwise.
  • the terms “installation”, “connected”, “connected”, “fixed” and the like shall be understood broadly, and may be either a fixed connection or a detachable connection, unless explicitly stated and defined otherwise. , or integrated; can be mechanical or electrical connection; can be directly connected, or indirectly connected through an intermediate medium, can be the internal communication of two elements or the interaction of two elements, unless otherwise specified Limited.
  • the specific meanings of the above terms in the present invention can be understood on a case-by-case basis.
  • the first feature "on” or “under” the second feature may be a direct contact of the first and second features, or the first and second features may be indirectly through an intermediate medium, unless otherwise explicitly stated and defined. contact.
  • the first feature "above”, “above” and “above” the second feature may be that the first feature is directly above or above the second feature, or merely that the first feature level is higher than the second feature.
  • the first feature “below”, “below” and “below” the second feature may be that the first feature is directly below or obliquely below the second feature, or merely that the first feature level is less than the second feature.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • General Physics & Mathematics (AREA)
  • Optics & Photonics (AREA)
  • Theoretical Computer Science (AREA)
  • Studio Devices (AREA)
  • Lenses (AREA)

Abstract

一种广角镜头的畸变矫正方法、装置（10）及系统，其中，方法包括：获取广角镜头的n个半视场角θ 1至θ n，并根据n个半视场角获取相邻视场比值β 1至β (n-1)（S101）；根据n个半视场角得到n个半视场角对应的像高r 1至r n（S102）；根据相邻视场比值β 1至β (n-1)和n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高（S103）；根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变（S104）。矫正方法使中心视场的采样率高于边缘，有效提高矫正效率，降低成本，简单易实现。

Description

广角镜头的畸变矫正方法、装置及系统
相关申请的交叉引用
本申请要求江西联创电子有限公司于2018年2月1日提交的、发明名称为“广角镜头的畸变矫正方法、装置及系统”的、中国专利申请号“201810102392.0”的优先权。
技术领域
本发明涉及图像处理技术领域，特别涉及一种广角镜头的畸变矫正方法、装置及系统。
背景技术
ADAS(Advanced Driver Assistance System，先进驾驶辅助系统)是利用安装在车上的各式各样的传感器，在第一时间收集车内外的环境数据，进行静、动态物体的辨识、侦测与追踪等技术上的处理，从而能够让驾驶者在最快的时间察觉到可能发生的危险，有效增加汽车驾驶的舒适性和安全性。因此，ADAS在实现自动驾驶的过程中扮演着非常重要的角色。
相关技术中，典型的ADAS系统一般在车辆前端设有三个摄像头，视场角度大概为42度、60度和120度，如果需要双目视觉，摄像头的数目还需要加倍；每个摄像头配备有一个2MP(Mega-Pixels，百万像素)，即200万像素的CMOS传感器。
然而，相关技术中的设置存在很大的冗余度，如42度相机本身的视场被60度相机和120度相机视场所覆盖，使得多个相机不但增加了系统连接的复杂度以及各个相机之间通信和同步的复杂度，而且使整个系统的成本随之升高，并且各个相机之间需要很好的保证视场的重叠，导致相互之间需要有一个固定的角度关系，同时整体和汽车本身也要保证角度，因此需要定期的做相机之间以及相机与车体之间的矫正，三个相机矫正的本身导致相对复杂且工作量大，亟待解决。
发明内容
本发明旨在至少在一定程度上解决相关技术中的技术问题之一。
为此，本发明的一个目的在于提出一种广角镜头的畸变矫正方法，该方法可以实现广角镜头的中心视场采样率高于边缘，从而可使用一个相机系统代替三个相机系统。
本发明的另一个目的在于提出一种广角镜头的畸变矫正装置。
本发明的再一个目的在于提出一种广角镜头的畸变矫正系统。
为达到上述目的，本发明一方面实施例提出了一种广角镜头的畸变矫正方法，包括以下步骤：获取广角镜头的n个半视场角θ 1至θ n，其中，θ n为θ (n-1)的相邻半视场角，并根据所述n个半视场角θ 1至θ n获取所述相邻视场比值β 1至β (n-1)，n为正整数；根据所述n个半视场角θ 1至θ n得到所述n个半视场角对应的像高r 1至r n；根据所述相邻视场比值β 1至β (n-1)和所述n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，并根据最大视场角对应的像高和所述相邻半视场角的像高关系通过递归计算得到所述每个半视场角的像高；根据所述每个半视场角的像高依次矫正所述广角镜头在中心视场的畸变。
本发明实施例的广角镜头的畸变矫正方法，可以根据相邻视场比值和多个半视场角对应的像高得到相邻半视场角的像高关系，并根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高，从而根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变，实现了广角镜头中心视场的采样率高于边缘，提高矫正的准确性和可靠性的同时，提高矫正效率，有效降低成本，简单易实现。
另外，根据本发明上述实施例的广角镜头的畸变矫正方法还可以具有以下附加的技术特征：
进一步地，在本发明的一个实施例中，所述获取广角镜头的n个半视场角θ 1至θ n，进一步包括：采集所述广角镜头的最大视场角；根据所述广角镜头的最大视场角得到所述半视场角θ 1，并根据相邻的前一项半视场角θ (n-1)得到半视场角θ n。
进一步地，在本发明的一个实施例中，所述根据所述相邻视场比值β 1至β (n-1)和所述n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，进一步包括：通过视场角关系式对所述n个半视场角对应的像高r 1至r n进行采样处理，其中，所述采样处理采用插值方法；根据所述视场角关系式得到相邻半视场角的像高关系式。
进一步地，在本发明的一个实施例中，所述相邻半视场角的像高关系式为：
Figure PCTCN2018081908-appb-000001
并且，所述根据最大视场角对应的像高和所述相邻半视场角的像高关系通过递归计算得到所述每个半视场角的像高：
Figure PCTCN2018081908-appb-000002
其中，r (n-1)为半视场角θ (n-1)对应的像高，r n为半视场角θ n对应的像高，β 1为θ 1和θ 2的 比值，β 2为θ 2和θ 3的比值，β (n-1)为θ (n-1)和θ n的比值。
进一步地，在本发明的一个实施例中，所述最大视场角大于等于70°。
进一步地，在本发明的一个实施例中，所述最大视场角大于等于100°且小于等于190°。
为达到上述目的，本发明另一方面实施例提出了一种广角镜头的畸变矫正装置，包括：采集模块，用于获取广角镜头的n个半视场角θ 1至θ n，其中，θ n为θ (n-1)的相邻半视场角，并根据所述n个半视场角θ 1至θ n获取所述相邻视场比值β 1至β (n-1)，n为正整数；获取模块，用于根据所述n个半视场角θ 1至θ n得到所述n个半视场角对应的像高r 1至r n；计算模块，用于根据所述相邻视场比值β 1至β (n-1)和所述n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，并根据最大视场角对应的像高和所述相邻半视场角的像高关系通过递归计算得到所述每个半视场角的像高；矫正模块，用于根据所述每个半视场角的像高依次矫正所述广角镜头在中心视场的畸变。
本发明实施例的广角镜头的畸变矫正装置，可以根据相邻视场比值和多个半视场角对应的像高得到相邻半视场角的像高关系，并根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高，从而根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变，实现了广角镜头中心视场的采样率高于边缘，提高矫正的准确性和可靠性的同时，提高矫正效率，有效降低成本，简单易实现。
另外，根据本发明上述实施例的广角镜头的畸变矫正装置还可以具有以下附加的技术特征：
进一步地，在本发明的一个实施例中，所述采集模块，进一步包括：采集单元，用于采集所述广角镜头的最大视场角；选取单元，用于根据所述广角镜头的最大视场角得到所述半视场角θ 1，并根据相邻的前一项半视场角θ (n-1)得到半视场角θ n。
进一步地，在本发明的一个实施例中，所述计算模块，进一步包括：处理单元，用于通过视场角关系式对所述n个半视场角对应的像高r 1至r n进行采样处理，其中，所述采样处理采用插值方法；获取单元，用于根据所述视场角关系式得到相邻半视场角的像高关系式。
进一步地，在本发明的一个实施例中，所述相邻半视场角的像高关系式为：
Figure PCTCN2018081908-appb-000003
并且，所述根据最大视场角对应的像高和所述相邻半视场角的像高关系通过递归计算得到所述每个半视场角的像高：
Figure PCTCN2018081908-appb-000004
其中，r (n-1)为半视场角θ (n-1)对应的像高，r n为半视场角θ n对应的像高，β 1为θ 1和θ 2的比值，β 2为θ 2和θ 3的比值，β (n-1)为θ (n-1)和θ n的比值。
进一步地，在本发明的一个实施例中，所述最大视场角大于或等于70°。
进一步地，在本发明的一个实施例中，所述最大视场角大于等于100°且小于等于190°。
为达到上述目的，本发明再一方面实施例提出一种广角镜头的畸变矫正系统，其包括上述的广角镜头的畸变矫正装置矫正过的广角镜头和像素传感器。
本发明实施例的广角镜头的畸变矫正系统，可以根据相邻视场比值和多个半视场角对应的像高得到相邻半视场角的像高关系，并根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高，从而根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变，实现了广角镜头中心视场的采样率高于边缘，提高矫正的准确性和可靠性的同时，提高矫正效率，有效降低成本，简单易实现。
进一步地，在本发明的一个实施例中，所述广角镜头从物侧到像侧依次包括第一透镜、第二透镜、第三透镜、光阑、第四透镜、第五透镜、第六透镜、第七透镜、滤光片。
进一步地，在本发明的一个实施例中，所述第一透镜、所述第二透镜和所述第七透镜均采用非球面透镜。
本发明附加的方面和优点将在下面的描述中部分给出，部分将从下面的描述中变得明显，或通过本发明的实践了解到。
附图说明
本发明上述的和/或附加的方面和优点从下面结合附图对实施例的描述中将变得明显和容易理解，其中：
图1为根据本发明实施例的广角镜头的畸变矫正方法的流程图；
图2为根据本发明一个实施例的广角镜头在相邻视场角的像高关系示意图；
图3为根据本发明实施例的广角镜头的畸变矫正装置的结构示意图；
图4为根据本发明一个实施例的采用递归畸变矫正的广角镜头的结构示意图；
图5为根据本发明一个具体实施例的广角镜头畸变矫正系统在降采样前的格点图示意图；以及
图6为根据本发明一个具体实施例的广角镜头畸变矫正系统在降采样后的格点图示意 图。
具体实施方式
下面详细描述本发明的实施例，所述实施例的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施例是示例性的，旨在用于解释本发明，而不能理解为对本发明的限制。
在介绍本发明实施例的广角镜头的畸变矫正方法、装置及系统之前，先简单介绍下相关技术中的广角镜头的畸变矫正方法。
市场上和ADAS系统摄像头配备的像素数为2MP的传感器本身尺寸约为1/3"，并且下一代即将推出的传感器，像素数为8MP。基于此，能否设计一款相机，使用一个高像素传感器相机而达到现有三个传感器相机的效果？如果简单的根据像素的叠加，那么最新的8MP传感器的像素数已经超过了三个2MP传感器的总和。然而，8MP的传感器可以满足最大视场角120°的像素要求，但是不一定满足中心视场角60°和42°的2MP的像素要求。
相关技术中，典型的成像光学系统存在靠近中心视场的采样率低、边缘采样率高的问题，同时边缘的照度也会由于视场角的加大而下降。如果成像系统按照理想畸变曲线矫正(r＝f*tanθ，θ为半视场角)，假设120度横向视场的相对像高为1，那么中心60度全视场所对应的相对像高为tan(30°)/tan(60°)≈0.333，中心42度全视场对应像高tan(21°)/tan(60°)≈0.222。如此以来，120度时8MP的图片，在截取了中心对应60度的区域后仅为0.9MP，对应于42度全视场更是下降到了0.4MP。这种理想畸变的矫正方法显然无法实现最初的要求。
类似的，如果系统按照完美的f-theta畸变矫正(r＝f*θ，θ为半视场角)方法，那么中心60度全视场所对应的相对像高为(30°)/(60°)＝0.5，中心42度全视场对应像高(21°)/(60°)＝0.35，于是120度时8MP的图片，在截取了中心对应60度区域后为2MP，在截取中心对应的42度视场后为1MP，f-theta畸变矫正方法，显然也无法达到最初的期望。
综合来看，相关技术中的理想畸变矫正方法和f-theta畸变矫正方法最主要的问题是边缘过高的采样率。对于理想畸变矫正而言，边缘采样率高于中心视场，即边缘使力过大；对于f-theta畸变矫正，则是边缘和中心采样率一致，即均匀使力。
本发明正是基于上述问题，而提出的一种广角镜头的畸变矫正方法及装置。
下面参照附图描述根据本发明实施例提出的广角镜头的畸变矫正方法、装置及系统，首先将参照附图描述根据本发明实施例提出的广角镜头的畸变矫正方法。
图1是本发明实施例的广角镜头的畸变矫正方法的流程图。
如图1所示，该广角镜头的畸变矫正方法包括以下步骤：
在步骤S101中，获取广角镜头的n个半视场角θ 1至θ n，其中，θ n为θ (n-1)的相邻半视场角，并根据n个半视场角θ 1至θ n获取相邻视场比值β 1至β (n-1)，n为正整数。
进一步地，在本发明的一个实施例中，获取广角镜头的n个半视场角θ 1至θ n，进一步包括：采集广角镜头的最大视场角；根据广角镜头的最大视场角得到半视场角θ 1，并根据相邻的前一项半视场角θ (n-1)得到半视场角θ n。
优选地，在本发明的一个实施例中，最大视场角大于或等于70°。
优选地，在本发明的一个实施例中，最大视场角大于等于100°且小于等于190°。
可以理解的是，广角镜头又称短焦距镜头，是视场角度大于60°的镜头；对于摄像机上的变焦镜头而言是焦距小于25mm的部分，广角镜头的使用有利于近距离表现大范围景物，并且广角镜头适于展现画面主体及其所处的环境，利用广角镜头景深的特点，可以对物体进行多层次的表现，增加画面的容量和信息量；利用广角镜头近距离接近拍摄对象，可以完成抢拍和偷拍；同时有利于维持移动摄像的平稳画面。
可以理解的是，本发明实施例可以首先获取广角镜头的最大HFOV(Horizontal Field Of View，视场角)，并依次选取n个广角镜头相邻的半视场角，如依次选取广角镜头相邻的半视场角θ 1，θ 2，...θ n，其中，θ 1＝HFOV/2，n>＝3，θ 1>θ 2>θ 3>...>θ n，需要说明的是，广角镜头的最大视场角可以大于或等于70°，优选的广角镜头的最大视场角大于等于100°且小于等于190°。
在步骤S102中，根据n个半视场角θ 1至θ n得到n个半视场角对应的像高r 1至r n。
在步骤S103中，根据相邻视场比值β 1至β (n-1)和n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，并根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高。
进一步地，在本发明的一个实施例中，根据相邻视场比值β 1至β (n-1)和n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，进一步包括：通过视场角关系式对n个半视场角对应的像高r 1至r n进行采样处理，其中，采样处理采用插值方法；根据视场角关系式得到相邻半视场角的像高关系式。
其中，在本发明的一个实施例中，相邻半视场角的像高关系式为：
Figure PCTCN2018081908-appb-000005
并且，根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个 半视场角的像高：
Figure PCTCN2018081908-appb-000006
其中，r (n-1)为半视场角θ (n-1)对应的像高，r n为半视场角θ n对应的像高，β 1为θ 1和θ 2的比值，β 2为θ 2和θ 3的比值，β (n-1)为θ (n-1)和θ n的比值。
可以理解的是，本发明实施例可以根据依次选取n个广角镜头相邻的半视场角θ 1，θ 2，...θ n，并定义n个半视场角θ 1，θ 2，...θ n的每个半视场角对应的像高r 1，r 2，...r n，其中，θ 1＝HFOV/2，n>＝3，θ 1>θ 2>θ 3>...>θ n，相邻半视场角的像高关系示意图如图2所示。
另外，本发明实施例可以定义相邻视场比值依次为：
β 1＝θ 1/θ 2，β 2＝θ 2/θ 3，...，β (n-1)＝θ (n-1)/θ n，
从而根据相邻半视场角的比值对n个半视场角的像高通过插值进行采样处理，以得到关系式，其中，对广角镜头的n个半视场角的像高通过插值进行采样处理可以通过以下关系式得到实现：
Figure PCTCN2018081908-appb-000007
Figure PCTCN2018081908-appb-000008
……
Figure PCTCN2018081908-appb-000009
其中，r 1为半视场角θ 1对应的像高，r 2为半视场角θ 2对应的像高，r 3为半视场角θ 3对应的像高，r (n-1)为半视场角θ (n-1)对应的像高，r n为半视场角θ n对应的像高，β 1为θ 1和θ 2的比值，β 2为θ 2和θ 3的比值，β (n-1)为θ (n-1)和θ n的比值，n为大于等于3的整数。
需要说明的是，关系式(3)的含义为高解析度的中心小视场θ n对应的像高r n，在相邻的大视场θ (n-1)时可以进行降采样，线性降采样的倍率为视场比值β (n-1)，而新加入的大视场像高(r (n-1)-r n)部分，理想状况下不需要重新采样，即不会浪费现有的像素分辨率，其结果是大视场θ (n-1)本身实现和相邻小视场θ n完全一致的像素分辨率，即降采样后的大视场像高等于r n。
另外，本发明实施例可以将上述关系式(1)至(3)进一步优化，获得相邻半视场角 的像高关系：
Figure PCTCN2018081908-appb-000010
Figure PCTCN2018081908-appb-000011
……
Figure PCTCN2018081908-appb-000012
也就是说，由于已知广角镜头最大半视场角对应的像高为r 1，从而本发明实施例可以通过递归计算像高关系：
Figure PCTCN2018081908-appb-000013
在步骤S104中，根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变。
可以理解的是，本发明实施例可以根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变。
举例而言，本发明实施例采用递归畸变矫正的广角镜头系统，并且优选为8MP的CMOS(Complementary Metal Oxide Semiconductor，互补金属氧化物半导体)传感器和最大视场角为120度的广角镜头组成的光学系统，可以实现有效替换三个镜头(视场角分别为120度、60度和42度)和三个2MP传感器的系统，也即要实现整个广角镜头的视场为120度，对应的解析度为8MP，中心60度和42度的小视场，同样可实现中心2MP的解像度；具体步骤如下：
首先，本发明实施例选取广角镜头的最大视场角120度的一半θ 1为60度，其半视场角对应的像高为r 1；相邻的中视场角60度的一半θ 2为30度，其半视场角对应的像高为r 2；中心小视场角42度的一半θ 3为21度，其半视场角对应的像高为r 3；从而得到相邻视场比值：
β 1＝θ 1/θ 2＝60/30＝2，β 2＝θ 2/θ 3＝30/21＝10/7；
其次，本发明实施例对半视场角θ 2和θ 3对应的像高分别进行采样处理，通过以下关系式处理：
Figure PCTCN2018081908-appb-000014
Figure PCTCN2018081908-appb-000015
将上述关系式进一步优化，以获得相邻半视场角的像高关系：
Figure PCTCN2018081908-appb-000016
Figure PCTCN2018081908-appb-000017
由于已知广角镜头最大半视场角对应的像高为r 1，通过递归计算像高关系：
Figure PCTCN2018081908-appb-000018
Figure PCTCN2018081908-appb-000019
从而根据求得的每个半视场角的像高依次对上述广角镜头进行畸变矫正设计。
根据关系式(12)和(13)，假设120度视场角对应的像高r 1为1，那相邻的60度视场角对应的归一化像高r 2＝2/3*r 1≈0.6667；在计算中心42度视场角的像高时，本发明实施例可以采取递归算法，对应的归一化像高r 1＝10/13*2/3*r 1＝20/39≈0.5128。
另外，本发明实施例提供的递归畸变矫正方法与相关技术中的常见畸变矫正方法带来的有益效果如表1所示。由于这种畸变要求是基于系统的非线性采样提出的，并不是以人眼的本身喜好而来，而是完全针对机器视觉，而且本发明实施例采用的这种采样的具体实现是插值，而插值点本身是非线性的分布，因此可以提前计算出插值位置从而提高速度。
表1 为本发明实施例的递归畸变控制与相关技术中的畸变控制的像高对比表。
表1
Figure PCTCN2018081908-appb-000020
Figure PCTCN2018081908-appb-000021
本发明实施例采用的是像素数为8MP的CMOS传感器及最大视场角为120度的广角镜头，如果将像高关系换算成像素关系，比如安森美8MP的AR0820，像素为3840*2160，那么半视场的像素为1920*1080，可以得到本发明实施例的递归畸变控制与相关技术中的畸变控制像素的对应关系，如表2所示。表2为本发明实施例的递归畸变控制与相关技术中的畸变控制的横向像素对比表。
表2
Figure PCTCN2018081908-appb-000022
从表1和表2可以看出，本发明实施例采用的递归畸变矫正控制的结果是可以保证广角镜头中心部分在截取后实现2MP的分辨率。具体而言，根据表2可知，如果只截取中心部分半视场角21度以内的信息，则对应像素数为985，和安森美2MP的AR0231的964(1928/2)像素基本一致。也就是说，这部分所包含的像素信息量和单独为之设计一个镜头并配备AR0231是一致的。
进一步地，对于半视场角30度，在对中心21度部分降采样后，所包含像素一样为985(985*7/10+1280-985＝985)。类似的，在60度视场情况下，对30度视场降采样以后的有效像素分辨同样为985。
综上，本发明实施例提供的采用非线性递归畸变矫正的广角镜头，120度视场角对应的解析度为8MP，截取中心42度的小视场，可以实现中心2MP的解像度；对中心42度视场做降采样，并结合相邻60度视场信息，同样可实现2MP的解像度；进一步对这个60度视场的图像降采样，结合60度和120度之间的信息，能够实现全视场的2MP解析度。因此，本发明实施例可以由一个更高像素的传感器和一个递归畸变矫正的广角镜头取代多个低像素的传感器和多个镜头系统。
由于中心部分采样率更高，在降采样的过程中，可以加入去噪声的算法，解决由于像素尺寸减小而引起的暗噪声(dark noise)增大的问题。比如中心部分像素的取均值(binning)。举例而言，如果是两倍的降采样，可以简单的先对中心部分做一个2×2的矩阵核(Kernel)卷积，然后采样。
一个典型的应用场景是白天光线充足时，中心部分比如42度视场角内解像力足够，信号强度足够，ADAS系统可以看的很远；晚上光线不足，由于车灯路灯的照度限制，42度的相机看的虽远，但是获得的照片噪声很大，照度不够，此时可以对60度视场进行降采样处理，去除部分噪声。此时虽然牺牲了42度以内中心部分的解像力，但是提高了系统的低光敏感度。由于夜间行车速度可以低一些，所需要辨别的物体解像力低些也是可以接受的。更进一步，如果照度更低，可见度更差，可以基于120度视场的情况，对中心的60度和42度同时去噪声和降采样。
换言之，本发明实施例可以用一个8MP的CMOS芯片，现在产生了三个2MP的图像，分别对应于42度视场、60度视场和120度视场。光线充足时可以同时用42度视场和120度视场服务ADAS系统，此时42度视场保证ADAS系统可以看的很远。光线稍弱时靠60度视场和120度视场。光线非常弱的时候可以只靠120度视场产生的2MP图像。此时8MP的传感器芯片，像素尺寸为2.1um，实现的效果相当于对传感器做了2×2的取均值(binning)，有效像素尺寸为4.2um，这个尺寸和低光表现非常好的AR0220芯片是一致的。
与相关技术中的畸变矫正控制方法相比，本发明实施例的广角镜头畸变矫正方法还具有以下有益效果：(1)本发明实施例有效矫正了广角镜头的畸变，实现了中心视场的采样率高于边缘，使广角镜头中心视场的解析度可以随系统需要实现定量调节；(2)由于很大度数的信息被压缩到了中心视场比较小的区域，使得镜头系统的相对照度可以甚至大于1，使得在光学设计时可以提高CRA(Chief Ray Angle，光线入射角)，从而减小系统的总长，使整个系统更紧凑。
以表2为例，在实际的光学镜头设计中，对畸变的控制可以加入几个关键点，具体需要控制的是1920像素和985像素处的像高，如需要可以加入1280像素处畸变点，保证畸变曲线在此三点中光滑过渡。
这里强调对1280像素处畸变无需严格控制，主要原因是此处接近中心视场，即21度半视场和30度半视场距离很近，对主光线的有效控制可能有一定挑战。换言之，即使30度的视场无法有效控制，其直接结果就是在对30度视场采样时，所得到的图片的大小超过2MP(比如3MP)，这种情况很容易由后续的数字处理进行进一步矫正；另外一个可以考虑的折衷是对30度半视场严格控制，而21度半视场本身就比较难做畸变矫正，其后果是21度半视场所切割出来的图片本身没有2MP。具体方案的选择，需要本领域技术人员根据实际情况进行决定，但是，60度处的畸变是实现这个算法的一个关键，此处畸变需要有效控制。
根据本发明实施例提出的广角镜头的畸变矫正方法，可以根据相邻视场比值和多个半 视场角对应的像高得到相邻半视场角的像高关系，并根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高，从而根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变，实现了广角镜头中心视场的采样率高于边缘，使广角镜头中心视场的解析度可以随系统需要实现定量调节，并且可以通过提高光线入射角以减小系统的总长，使整个系统更紧凑；通过一个高像素相机有效取代多个低像素相机，节省ADAS系统成本的同时简化了系统矫正和系统计算方面的要求，提高矫正的准确性和可靠性的同时，提高矫正效率，有效降低成本，简单易实现。
其次参照附图描述根据本发明实施例提出的广角镜头的畸变矫正装置。
图3是本发明实施例的广角镜头的畸变矫正装置的结构示意图。
如图3所示，该广角镜头的畸变矫正装置10包括：采集模块100、获取模块200、计算模块300和矫正模块400。
其中，采集模块100用于获取广角镜头的n个半视场角θ 1至θ n，其中，θ n为θ (n-1)的相邻半视场角，并根据n个半视场角θ 1至θ n获取相邻视场比值β 1至β (n-1)，n为正整数。获取模块200用于根据n个半视场角θ 1至θ n得到n个半视场角对应的像高r 1至r n。计算模块300用于根据相邻视场比值β 1至β (n-1)和n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，并根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高。矫正模块400用于根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变。本发明实施例的装置10可以根据相邻半视场角的像高关系矫正广角镜头在中心视场的畸变，使得中心视场的采样率高于边缘，降低成本，简单易实现。
进一步地，在本发明的一个实施例中，采集模块100进一步包括：采集单元和选取单元。其中，采集单元用于采集广角镜头的最大视场角。选取单元用于根据广角镜头的最大视场角得到半视场角θ 1，并根据相邻的前一项半视场角θ (n-1)得到半视场角θ n。
进一步地，在本发明的一个实施例中，计算模块300进一步包括：处理单元和计算单元。其中，处理单元用于通过视场角关系式对n个半视场角对应的像高r 1至r n进行采样处理，其中，采样处理采用插值方法。获取单元用于根据视场角关系式得到相邻半视场角的像高关系式。
进一步地，在本发明的一个实施例中，相邻半视场角的像高关系式为：
Figure PCTCN2018081908-appb-000023
并且，根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个 半视场角的像高：
Figure PCTCN2018081908-appb-000024
其中，r (n-1)为半视场角θ (n-1)对应的像高，r n为半视场角θ n对应的像高，β 1为θ 1和θ 2的比值，β 2为θ 2和θ 3的比值，β (n-1)为θ (n-1)和θ n的比值。
进一步地，在本发明的一个实施例中，最大视场角大于或等于70°。
进一步地，在本发明的一个实施例中，最大视场角大于等于100°且小于等于190°。
需要说明的是，前述对广角镜头的畸变矫正方法实施例的解释说明也适用于该实施例的广角镜头的畸变矫正装置，此处不再赘述。
根据本发明实施例提出的广角镜头的畸变矫正装置，可以根据相邻视场比值和多个半视场角对应的像高得到相邻半视场角的像高关系，并根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高，从而根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变，实现了广角镜头中心视场的采样率高于边缘，使广角镜头中心视场的解析度可以随系统需要实现定量调节，并且可以通过提高光线入射角以减小系统的总长，使整个系统更紧凑。
此外，本发明实施例还提出了一种广角镜头的畸变矫正系统，该系统包括上述的广角镜头的畸变矫正装置矫正过的广角镜头和像素传感器。
进一步地，在本发明的一个实施例中，采用上述畸变矫正装置矫正过的广角镜头从物侧到像侧依次包括第一透镜、第二透镜、第三透镜、光阑、第四透镜、第五透镜、第六透镜、第七透镜、滤光片。
进一步地，在本发明的一个实施例中，第一透镜、第二透镜和第七透镜均采用非球面透镜。
进一步地，在本发明的一个实施例中，所述第一透镜、所述第二透镜、所述第七透镜的非球面表面形状均满足下列方程：
Figure PCTCN2018081908-appb-000025
其中，z为半径所对应的曲率，h为径向坐标，c为曲面顶点的曲率，K为圆锥二次曲线系数，B、C、D、E分别表示四阶、六阶、八阶、十阶径向坐标所对应的系数。
可以理解的是，如图4所示，本发明实施例采用的递归畸变矫正的广角镜头(最大视场角为120度)从物侧到成像面依次包括：第一透镜L1、第二透镜L2、第三透镜L3、光 阑S1、第四透镜L4、第五透镜L5、第六透镜L6、第七透镜L7、滤光片S2，并且第一透镜L1、第二透镜L2、第七透镜L7为非球面透镜。本发明实施例的广角镜头中各个镜片的设计参数如表3所示，各透镜的非球面参数如表4所示。
表3
Figure PCTCN2018081908-appb-000026
Figure PCTCN2018081908-appb-000027
表4
Figure PCTCN2018081908-appb-000028
Figure PCTCN2018081908-appb-000029
进一步地，如图5所示，小x是主光线对视场的一个均匀采样，在投影到了CMOS传感器后，由于畸变的存在，整体发生了形变，结果是中心部分没有太大畸变，一个小x可以对应大概9个像素的采样，因此分辨率高，而边上的一个小x只对应一个像素的采样，分辨率低，图中的黑色小方格代表一个像素。在经过降采样后，这些小x又会均匀的分布，重建的图像如图6所示。这个重新采样的频率，可以计算为((30/21)*(60/30))^2＝8.2。因此，本发明实施例提供的广角镜头，在重新采样后同时实现了畸变的矫正。
表5 是本实施例提供的广角镜头在不同半视场角时对应的像高及像素参数，从表5中可以看出，本发明实施例的广角镜头和本发明递归畸变矫正方法的理论值基本一致，说明本发明实施例的递归畸变矫正方法能够有效矫正广角镜头在中心视场的畸变，能够实现中心视场的采样率高于边缘的采样率。
表5
  θ 3＝42/2＝21度 θ 2＝60/2＝30度 θ 1＝120/2＝60度
递归畸变控制像高 0.51 0.67 1.00
实际设计像高(mm) 2.00 2.65 4.03
实际设计像高归一化 0.50 0.66 1.00
对应水平像素(Pixels) 1904.8 2523.8 3838.1
对应视角像素(Pixels) 2,040,816 3,582,908 8,286,173
举例而言，在本发明的另一个实施例中，通过采用递归畸变矫正的广角镜头，并且假定采用一个160度的广角镜头和12.4MP的CMOS传感器的系统，如果依然保持中心42度视场的解析度为2MP，那么根据表6可知，如果增加横向视场到160度以后，传感器需要的像素为4802*2702，即12.4MP。另外如果保持横向视场120度不变，在传感器增加到12MP以后，中心42度的视场的像素数目也增加到3MP。因此，下一代广角镜头产品的发展可以有两个方向，一是提高视场覆盖，保持中心解像度；二是保持视场覆盖，提高中心解像度。 需要说明的是，本发明实施例可以根据ADAS的发展和算法决定更适合的方案。
表6
Figure PCTCN2018081908-appb-000030
根据表6可知，在HFOV达到160度的时候，如果需要保证中心42度有2MP的覆盖，那么整个传感器像素为4802*2702(即12.4MP)。
需要说明的是，前述对广角镜头的畸变矫正方法实施例的解释说明也适用于该实施例的广角镜头的畸变矫正系统，此处不再赘述。
根据本发明实施例提出的广角镜头的畸变矫正系统，可以根据相邻视场比值和多个半视场角对应的像高得到相邻半视场角的像高关系，并根据最大视场角对应的像高和相邻半视场角的像高关系通过递归计算得到每个半视场角的像高，从而根据每个半视场角的像高依次矫正广角镜头在中心视场的畸变，实现了广角镜头中心视场的采样率高于边缘，使广角镜头中心视场的解析度可以随系统需要实现定量调节，并且可以通过提高光线入射角以减小系统的总长，使整个系统更紧凑；通过一个高像素相机有效取代多个低像素相机，节省ADAS系统成本的同时简化了系统矫正和系统计算方面的要求，提高矫正的准确性和可靠性的同时，提高矫正效率，有效降低成本，简单易实现。
在本发明的描述中，需要理解的是，术语“中心”、“纵向”、“横向”、“长度”、“宽度”、“厚度”、“上”、“下”、“前”、“后”、“左”、“右”、“竖直”、“水平”、“顶”、“底”“内”、“外”、“顺时针”、“逆时针”、“轴向”、“径向”、“周向”等指示的方位或位置关系为基于附图所示的方位或位置关系，仅是为了便于描述本发明和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本发明的限制。
此外，术语“第一”、“第二”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括至少一个该特征。在本发明的描述中，“多个”的含义是至少两个，例如两个，三个等，除非另有明确具体的限定。
在本发明中，除非另有明确的规定和限定，术语“安装”、“相连”、“连接”、“固定”等术 语应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或成一体；可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连，可以是两个元件内部的连通或两个元件的相互作用关系，除非另有明确的限定。对于本领域的普通技术人员而言，可以根据具体情况理解上述术语在本发明中的具体含义。
在本发明中，除非另有明确的规定和限定，第一特征在第二特征“上”或“下”可以是第一和第二特征直接接触，或第一和第二特征通过中间媒介间接接触。而且，第一特征在第二特征“之上”、“上方”和“上面”可是第一特征在第二特征正上方或斜上方，或仅仅表示第一特征水平高度高于第二特征。第一特征在第二特征“之下”、“下方”和“下面”可以是第一特征在第二特征正下方或斜下方，或仅仅表示第一特征水平高度小于第二特征。
在本说明书的描述中，参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本发明的至少一个实施例或示例中。在本说明书中，对上述术语的示意性表述不必须针对的是相同的实施例或示例。而且，描述的具体特征、结构、材料或者特点可以在任一个或多个实施例或示例中以合适的方式结合。此外，在不相互矛盾的情况下，本领域的技术人员可以将本说明书中描述的不同实施例或示例以及不同实施例或示例的特征进行结合和组合。
尽管上面已经示出和描述了本发明的实施例，可以理解的是，上述实施例是示例性的，不能理解为对本发明的限制，本领域的普通技术人员在本发明的范围内可以对上述实施例进行变化、修改、替换和变型。

Claims (15)

  1. 一种广角镜头的畸变矫正方法，其特征在于，包括以下步骤：
    获取广角镜头的n个半视场角θ 1至θ n，其中，θ n为θ (n-1)的相邻半视场角，并根据所述n个半视场角θ 1至θ n获取所述相邻视场比值β 1至β (n-1)，n为正整数；
    根据所述n个半视场角θ 1至θ n得到所述n个半视场角对应的像高r 1至r n；
    根据所述相邻视场比值β 1至β (n-1)和所述n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，并根据最大视场角对应的像高和所述相邻半视场角的像高关系通过递归计算得到所述每个半视场角的像高；以及
    根据所述每个半视场角的像高依次矫正所述广角镜头在中心视场的畸变。
  2. 根据权利要求1所述的广角镜头的畸变矫正方法，其特征在于，所述获取广角镜头的n个半视场角θ 1至θ n，进一步包括：
    采集所述广角镜头的最大视场角；
    根据所述广角镜头的最大视场角得到所述半视场角θ 1，并根据相邻的前一项半视场角θ (n-1)得到半视场角θ n。
  3. 根据权利要求1所述的广角镜头的畸变矫正方法，其特征在于，所述根据所述相邻视场比值β 1至β (n-1)和所述n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，进一步包括：
    通过视场角关系式对所述n个半视场角对应的像高r 1至r n进行采样处理，其中，所述采样处理采用插值方法；
    根据所述视场角关系式得到相邻半视场角的像高关系式。
  4. 根据权利要求3所述的广角镜头的畸变矫正方法，其特征在于，所述相邻半视场角的像高关系式为：
    Figure PCTCN2018081908-appb-100001
    并且，所述根据最大视场角对应的像高和所述相邻半视场角的像高关系通过递归计算得到所述每个半视场角的像高：
    Figure PCTCN2018081908-appb-100002
    其中，r (n-1)为半视场角θ (n-1)对应的像高，r n为半视场角θ n对应的像高，β 1为θ 1和θ 2的 比值，β 2为θ 2和θ 3的比值，β (n-1)为θ (n-1)和θ n的比值。
  5. 根据权利要求2所述的广角镜头的畸变矫正方法，其特征在于，所述最大视场角大于或等于70°。
  6. 根据权利要求5所述的广角镜头的畸变矫正方法，其特征在于，所述最大视场角大于等于100°且小于等于190°。
  7. 一种广角镜头的畸变矫正装置，其特征在于，包括：
    采集模块，用于获取广角镜头的n个半视场角θ 1至θ n，其中，θ n为θ (n-1)的相邻半视场角，并根据所述n个半视场角θ 1至θ n获取所述相邻视场比值β 1至β (n-1)，n为正整数；
    获取模块，用于根据所述n个半视场角θ 1至θ n得到所述n个半视场角对应的像高r 1至r n；
    计算模块，用于根据所述相邻视场比值β 1至β (n-1)和所述n个半视场角对应的像高r 1至r n得到相邻半视场角的像高关系，并根据最大视场角对应的像高和所述相邻半视场角的像高关系通过递归计算得到所述每个半视场角的像高；以及
    矫正模块，用于根据所述每个半视场角的像高依次矫正所述广角镜头在中心视场的畸变。
  8. 根据权利要求7所述的广角镜头的畸变矫正装置，其特征在于，所述采集模块，进一步包括：
    采集单元，用于采集所述广角镜头的最大视场角；
    选取单元，用于根据所述广角镜头的最大视场角得到所述半视场角θ 1，并根据相邻的前一项半视场角θ (n-1)得到半视场角θ n。
  9. 根据权利要求7所述的广角镜头的畸变矫正装置，其特征在于，所述计算模块，进一步包括：
    处理单元，用于通过视场角关系式对所述n个半视场角对应的像高r 1至r n进行采样处理，其中，所述采样处理采用插值方法；
    获取单元，用于根据所述视场角关系式得到相邻半视场角的像高关系式。
  10. 根据权利要求9所述的广角镜头的畸变矫正装置，其特征在于，所述相邻半视场角的像高关系式为：
    Figure PCTCN2018081908-appb-100003
    并且，所述根据最大视场角对应的像高和所述相邻半视场角的像高关系通过递归计算 得到所述每个半视场角的像高：
    Figure PCTCN2018081908-appb-100004
    其中，r (n-1)为半视场角θ (n-1)对应的像高，r n为半视场角θ n对应的像高，β 1为θ 1和θ 2的比值，β 2为θ 2和θ 3的比值，β (n-1)为θ (n-1)和θ n的比值。
  11. 根据权利要求8所述的广角镜头的畸变矫正装置，其特征在于，所述最大视场角大于或等于70°。
  12. 根据权利要求11所述的广角镜头的畸变矫正装置，其特征在于，所述最大视场角大于等于100°且小于等于190°。
  13. 一种广角镜头的畸变矫正系统，其特征在于，包括：
    如权利要求7-12任一项所述的广角镜头的畸变矫正装置矫正过的广角镜头和像素传感器。
  14. 根据权利要求13所述的广角镜头的畸变矫正系统，其特征在于，所述广角镜头的畸变矫正装置矫正过的广角镜头从物侧到像侧依次包括第一透镜、第二透镜、第三透镜、光阑、第四透镜、第五透镜、第六透镜、第七透镜和滤光片。
  15. 根据权利要求14所述的广角镜头的畸变矫正系统，其特征在于，所述第一透镜、所述第二透镜和所述第七透镜均采用非球面透镜。
PCT/CN2018/081908 2018-02-01 2018-04-04 广角镜头的畸变矫正方法、装置及系统 WO2019148642A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/195,284 US10748258B2 (en) 2018-02-01 2018-11-19 Method, device and system for correcting distortion of wide-angle lens

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810102392.0A CN108495066B (zh) 2018-02-01 2018-02-01 广角镜头的畸变矫正方法、装置及系统
CN201810102392.0 2018-02-01

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/195,284 Continuation US10748258B2 (en) 2018-02-01 2018-11-19 Method, device and system for correcting distortion of wide-angle lens

Publications (1)

Publication Number Publication Date
WO2019148642A1 true WO2019148642A1 (zh) 2019-08-08

Family

ID=63344390

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/081908 WO2019148642A1 (zh) 2018-02-01 2018-04-04 广角镜头的畸变矫正方法、装置及系统

Country Status (2)

Country Link
CN (1) CN108495066B (zh)
WO (1) WO2019148642A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112308792A (zh) * 2020-10-15 2021-02-02 深圳一块互动网络技术有限公司 预设参数的图像输出方法、终端、电子设备及存储介质

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109544484B (zh) * 2019-02-20 2019-06-07 上海赫千电子科技有限公司 一种图像校正方法和装置
CN110519486B (zh) * 2019-09-19 2021-09-03 Oppo广东移动通信有限公司 基于广角镜头的畸变补偿方法、装置、及相关设备
CN113747002B (zh) * 2020-05-29 2023-04-07 青岛海信移动通信技术股份有限公司 终端及图像拍摄方法
CN111861932B (zh) * 2020-07-28 2022-05-17 RealMe重庆移动通信有限公司 图像的畸变校正方法、装置及移动终端

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6707020B1 (en) * 1999-12-28 2004-03-16 Mza Associates Corporation Adaptive dynamic range wavefront sensor
KR100871044B1 (ko) * 2007-09-28 2008-11-27 한국오므론전장주식회사 차량 후방안내 시스템의 온스크린 디스플레이 라인 사전생성 방법 및 이를 출력하는 차량 후방이동 궤적 안내방법
CN101572828A (zh) * 2009-05-20 2009-11-04 长春理工大学 基于gpu的相机/摄像机实时畸变校正方法
CN101582164A (zh) * 2009-06-24 2009-11-18 北京锦恒佳晖汽车电子系统有限公司 一种辅助倒车系统的图像处理方法
CN102156969A (zh) * 2011-04-12 2011-08-17 潘林岭 图像纠偏处理方法
CN105827899A (zh) * 2015-05-26 2016-08-03 维沃移动通信有限公司 一种修正镜头畸变的方法和装置

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101432650A (zh) * 2006-04-10 2009-05-13 阿里克斯·宁 超广角成像物镜
JP5275718B2 (ja) * 2008-08-05 2013-08-28 オリンパス株式会社 結像光学系およびそれを有する電子撮像装置
US8000025B2 (en) * 2009-01-08 2011-08-16 Panasonic Corporation Zoom lens system, imaging device and camera
CN103793113B (zh) * 2014-03-10 2019-04-16 天津津航技术物理研究所 一种光学触摸模组的成像定位方法及光学触摸控制设备
CN204536640U (zh) * 2015-04-10 2015-08-05 中山联合光电科技股份有限公司 一种新型的红外共焦鱼眼镜头
CN106324813B (zh) * 2015-06-26 2019-03-05 信泰光学(深圳)有限公司 变焦镜头
CN106358034A (zh) * 2016-10-19 2017-01-25 深圳市麦极客图像技术有限公司 录制、观看vr视频的装置、设备及vr视频录放系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6707020B1 (en) * 1999-12-28 2004-03-16 Mza Associates Corporation Adaptive dynamic range wavefront sensor
KR100871044B1 (ko) * 2007-09-28 2008-11-27 한국오므론전장주식회사 차량 후방안내 시스템의 온스크린 디스플레이 라인 사전생성 방법 및 이를 출력하는 차량 후방이동 궤적 안내방법
CN101572828A (zh) * 2009-05-20 2009-11-04 长春理工大学 基于gpu的相机/摄像机实时畸变校正方法
CN101582164A (zh) * 2009-06-24 2009-11-18 北京锦恒佳晖汽车电子系统有限公司 一种辅助倒车系统的图像处理方法
CN102156969A (zh) * 2011-04-12 2011-08-17 潘林岭 图像纠偏处理方法
CN105827899A (zh) * 2015-05-26 2016-08-03 维沃移动通信有限公司 一种修正镜头畸变的方法和装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112308792A (zh) * 2020-10-15 2021-02-02 深圳一块互动网络技术有限公司 预设参数的图像输出方法、终端、电子设备及存储介质

Also Published As

Publication number Publication date
CN108495066A (zh) 2018-09-04
CN108495066B (zh) 2020-01-31

Similar Documents

Publication Publication Date Title
WO2019148642A1 (zh) 广角镜头的畸变矫正方法、装置及系统
JP4885179B2 (ja) 映像歪曲補正方法及びその補正方法を採用した映像処理装置
US9686479B2 (en) Method for combining multiple image fields
US20110216157A1 (en) Object Detection and Rendering for Wide Field of View (WFOV) Image Acquisition Systems
EP2061234A1 (en) Imaging apparatus
JP2016001453A (ja) レンズ歪曲補正装置及びこれを含むアプリケーションプロセッサ
US11895403B2 (en) Imaging support device that performs registration control for setting a detected subject image position, imaging support system, imaging system, imaging support method, and program
WO2007129446A1 (ja) 画像処理方法、画像処理プログラム、画像処理装置、及び撮像装置
US20140022440A1 (en) Focus extending optical system and edof imaging system
WO2021179605A1 (zh) 基于gpu的摄像头视频投影方法、装置、设备及存储介质
CN106170065A (zh) 一种鱼眼摄像头的lsc补偿方法及装置
WO2012026502A1 (ja) 立体撮影装置および立体撮影方法
TWI663421B (zh) 廣角鏡頭的畸變矯正方法、裝置及系統
CN113608709A (zh) 显示屏校正方法、装置、可读存储介质及系统
CN111988517B (zh) 图像处理装置、镜头装置和图像处理方法
WO2014156712A1 (ja) 複眼光学系及び撮像装置
US10748258B2 (en) Method, device and system for correcting distortion of wide-angle lens
WO2019085023A1 (zh) 改善显示画面中鬼影的方法和系统
CN206864472U (zh) 图像传感器
US9832377B2 (en) Data acquiring method and electronic device thereof
JP2000341568A (ja) 撮像装置、撮像システム及び撮像方法
JP2016040883A (ja) 画像処理装置、画像処理方法、画像処理システム、及びプログラム
WO2013031564A1 (ja) 光学系及び撮像装置
JP3805631B2 (ja) 多重焦点全方位撮像装置
WO2022174438A1 (zh) 光学系统、摄像模组及电子设备

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: 18903949

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: 18903949

Country of ref document: EP

Kind code of ref document: A1