Hyperspectral optical system and automatic focusing method using same
By dividing the hyperspectral optical system into multiple separate regions, calculating representative depths of field, and using a hyperspectral CCD to capture and correct images, the problem of focus blurring in hyperspectral optical devices under non-uniform depths of field is solved, achieving clearer image capture.
Patent Information
- Application Number
- CN202580003647.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-05-28
- Filing Date
- 2025-05-19
- Publication Date
- 2026-02-24
AI Technical Summary
Existing hyperspectral optical equipment is prone to blurring when shooting targets that are uneven and whose depth of field is not uniform along the Z-axis.
By dividing the target into multiple separate areas, capturing DOF-specific images using the viewfinder imaging unit, calculating the representative depth of field for each area, capturing images of the area corresponding to the representative DOF using a hyperspectral CCD, and combining the images using an image correction unit to correct focus blur.
Even if the DOF of the target is uneven in the thickness direction, the autofocus technology can provide improved focus sharpness and prevent image blur.
Smart Images
Figure CN121569172A_ABST
Abstract
Description
[0001] Cross-reference to related applications
[0002] This application claims the benefit of priority based on Korean Patent Application No. 10-2024-0069400, filed on May 28, 2024, the entire contents of which are incorporated herein by reference as a part of this specification. Technical Field
[0003] One aspect of this disclosure relates to a hyperspectral optical system and an autofocusing method using the system. Background Technology
[0004] "Hyperspectral" refers to a technique that divides light into approximately 200 wavelengths in the visible and near-infrared regions, unlike general images that only distinguish the three primary colors of light. The human eye primarily perceives three bands of visible light, while hyperspectral imaging technology can divide the spectrum into even more bands.
[0005] Therefore, hyperspectral optical devices with hyperspectral imaging technology, such as hyperspectral cameras or hyperspectral sensors, have a large number of detection bands, and each band is further subdivided into narrower bands, allowing for the extraction of more information from the images captured using them. Recently, research utilizing hyperspectral imaging technology has been continuously deepening.
[0006] However, the problem is that when the target object is uneven and the depth of field (DOF) is not uniform along the Z-axis, images captured by currently commercially available hyperspectral optics exhibit blurred focus. The Z-axis refers to the depth of field (DOF) or thickness direction.
[0007] Therefore, there is a need to develop a technology that can prevent focus blur in images captured by hyperspectral optics, even when the target being photographed is uneven and the DOF is not uniform in the Z-axis direction.
[0008] [Existing Technical Documents]
[0009] [Patent Literature]
[0010] (Patent Document 1) US Patent Publication No. 2020-0200606 Summary of the Invention
[0011] Technical issues
[0012] In order to solve the above-mentioned problems, the inventors of the present invention have conducted various studies and found that when an autofocus technique is introduced that selects multiple depths of field (DOF) for the target of the hyperspectral optical system and then combines the images captured at each depth of field (DOF) to provide correction, the focus blur caused by the uneven depth of field of the target in the image captured by the hyperspectral optical system can be improved.
[0013] Therefore, one aspect of this disclosure is to provide a hyperspectral optical system capable of providing images with improved focus blur.
[0014] Another object of one aspect of this disclosure is to provide an autofocusing method using a hyperspectral optical system.
[0015] Technical solution
[0016] To achieve the above objectives, one aspect of this disclosure provides a hyperspectral optical system, comprising:
[0017] The viewfinder imaging unit divides the subject into multiple individual regions and then uses the viewfinder to capture a depth-of-field (DOF) specific image of the subject, including the individual regions.
[0018] (b) A representative depth of field (DOF) derivation unit for a single area of the target being captured, which calculates the focus of each of the captured DOF-specific viewfinder images and selects the DOF corresponding to the viewfinder image with good focus as the representative DOF of each single area;
[0019] (c) A hyperspectral image acquisition unit that uses a hyperspectral CCD (charge-coupled device) to capture individual regions corresponding to selected representative DOFs; and
[0020] (d) Image correction unit, which combines hyperspectral images of individual regions corresponding to representative DOF captured by hyperspectral CCD to correct focus blur caused by non-uniform DOF of the photographed target.
[0021] In one embodiment of this disclosure, the target to be photographed can be divided into 5 to 20 separate regions.
[0022] In one embodiment of this disclosure, 3 to 10 DOF-specific images can be captured for each individual region.
[0023] In one embodiment of this disclosure, the focus can be calculated in the viewfinder in a representative DOF deriving unit (b) for a separate area of the target to be photographed.
[0024] One aspect of this disclosure also provides an autofocus method using a hyperspectral optical system, comprising: (A) a viewfinder imaging step which divides the subject into multiple individual regions and then uses the viewfinder to capture depth-of-field (DOF) specific images of the individual regions;
[0025] (B) A representative depth of field (DOF) derivation step for individual regions of the target, which calculates the focus of each in the DOF-specific image of each individual region captured, and selects the DOF corresponding to the calculated focus as the representative DOF of each individual region;
[0026] (C) A hyperspectral image acquisition step, which repeatedly captures individual regions corresponding to selected representative DOFs using a hyperspectral CCD (charge-coupled device); and
[0027] (D) Image correction step, which combines hyperspectral images of individual regions corresponding to representative DOF captured by hyperspectral CCD to correct focus blur caused by non-uniform DOF of the photographed target.
[0028] In one embodiment of this disclosure, the target to be photographed can be divided into 5 to 20 separate regions.
[0029] In one embodiment of this disclosure, 3 to 10 DOF-specific images can be captured for each individual region.
[0030] In one embodiment of this disclosure, in step (B), the focus can be calculated in the viewfinder.
[0031] Invention Effects
[0032] According to one aspect of this disclosure, even if the DOF of the target image captured by the hyperspectral optical system is non-uniform in the thickness or depth direction, it is possible to provide an image with improved focus blur through autofocus technology. Attached Figure Description
[0033] Figure 1a This is a schematic diagram illustrating how the viewfinder imaging unit of a hyperspectral optical system uses the viewfinder to capture a single area of the subject. Figure 1b This shows depth-of-field (DOF) specific viewfinder image data for a single area captured by the viewfinder.
[0034] Figure 2a The division of specific individual regions of the photographed target using DOF is shown, and Figure 2b A graph showing the representative depth of field (DOF) of a single region selected by focusing the viewfinder image of a single region at a specific depth.
[0035] Figure 3aThis is a schematic diagram illustrating the repeated capture of individual regions using a hyperspectral CCD based on a representative DOF path. Figure 3b The image shows a hyperspectral CCD image (scanned image) captured via the path.
[0036] Figure 4a The process of combining the corrected images for each path is shown, and Figure 4b The corrected hyperspectral CCD image obtained by combining the corrected images of each path is shown.
[0037] Figure 5 An image captured by a hyperspectral optical system employing autofocus technology according to a preferred embodiment of the present invention is shown.
[0038] Figure 6 The image shown is captured by a conventional hyperspectral optical system without the application of autofocus technology. Detailed Implementation
[0039] The invention will now be described in more detail to aid in understanding it.
[0040] The terms or words used in this specification and claims should not be construed as limited to their ordinary or dictionary meanings, but should be interpreted based on the principle that the inventor can appropriately define the concepts of the terms to best explain his invention, in accordance with the technical concept of the invention.
[0041] The term "hyperspectral optical system" as used in this specification refers to an optical system that includes a hyperspectral CCD (charge-coupled device), and a hyperspectral optical system can also be understood as a hyperspectral CCD with autofocus functionality.
[0042] As used in this specification, the term "auto focusing" refers to a technique that selects multiple depths of field (DOF) for a target image of a hyperspectral optical device, captures an image of the target corresponding to each of these depths, and then combines them to improve the blurring caused by DOF differences.
[0043] The term "Z-axis" as used in this specification can refer to the depth or thickness direction of any object, and specifically, to the depth or thickness direction of the target being photographed.
[0044] Hyperspectral optical equipment
[0045] One aspect of this disclosure relates to hyperspectral optical devices.
[0046] According to one aspect of this disclosure, a hyperspectral optical device includes: (a) a viewfinder imaging unit that divides a target into multiple individual regions and then uses the viewfinder to capture a DOF-specific image of the target including the individual regions;
[0047] (b) A representative depth of field (DOF) derivation unit for a single area of the target being captured, which calculates the focus of each of the captured DOF-specific viewfinder images and selects the DOF corresponding to the viewfinder image with good focus as the representative DOF of each single area;
[0048] (c) A hyperspectral image acquisition unit that uses a hyperspectral CCD (charge-coupled device) to capture individual regions corresponding to selected representative DOFs; and
[0049] (d) Image correction unit, which combines hyperspectral images of individual regions corresponding to representative DOF captured by hyperspectral CCD to correct focus blur caused by non-uniform DOF of the photographed target.
[0050] By performing autofocus according to (a) to (d) above, focus blurring in images captured by hyperspectral optical systems can be prevented.
[0051] In a hyperspectral optical system of one aspect of this disclosure, (a) a viewfinder imaging unit can divide the target into multiple individual regions and then use the viewfinder to capture a DOF-specific image of the target including the individual regions.
[0052] Figure 1a This is a schematic diagram illustrating how the viewfinder imaging unit of a hyperspectral optical system uses the viewfinder to capture a single area of the subject. Figure 1b This shows depth-of-field (DOF) viewfinder-specific image data for a single area captured by the viewfinder.
[0053] refer to Figure 1a and Figure 1b After dividing the target 20 into multiple individual regions 21, the viewfinder 10 can be used to capture the target 20 including the multiple individual regions 21. The depth of field (DOF) may vary for each individual region 21 in the target 20. A viewfinder image 22 including each individual region 21 of the target 20 can be captured, and by capturing the target 20 including the individual regions 21 at multiple depths of field (DOF), a DOF-specific viewfinder image 22 of each individual region 21 can be obtained. For example, by dividing the DOF into 1 to n, viewfinder images from the first DOF viewfinder image 22 (1) to the nth DOF viewfinder image 22 (n) can be obtained. Thus, cumulative viewfinder image data accumulating the first to nth specific DOF viewfinder images 22 (1) to 22 (n) can be obtained.
[0054] A viewfinder is a device that allows the photographer to check the camera's line of sight, rather than the photographer's. It's a camera device used to frame the image to be taken and to focus more accurately on the subject. Typically, a viewfinder can be used for measuring position (X / Y axis) for exploration purposes, or to calculate the focus of an image captured at a specific depth of field (Z axis).
[0055] The viewfinder is not particularly restricted, as long as it is conventionally used in the field, and for example, optical viewfinders and electronic viewfinders can be used.
[0056] Optical viewfinders are called OVF (Optical View Finder), while electronic viewfinders are called EVF (Electronic View Finder). The difference between OVF and EVF can be said to be the type of lens used.
[0057] An optical viewfinder (OVF) is designed to use a prism to reflect light that enters through a lens, allowing the user to see the image directly. SLM film cameras allow photographers to preview captured images through the OVF and press the shutter button to take the picture. Because the image from the light is transmitted directly through the OVF, it has the advantage of delivering a sharper image—like the image of the image itself.
[0058] Compared to an OVF (Optical Viewfinder), an Electronic Viewfinder (EVF) improves camera portability by reducing the space occupied by the viewfinder device. An EVF electronically processes the light passing through the lens and reaching the image sensor, then processes this information and displays the processed image on a small screen. This screen can be viewed through the viewfinder.
[0059] In the hyperspectral optical system of one aspect of this disclosure, in (b) the representative depth of field (DOF) deriving unit for a single area of the target, the focus of each of the captured DOF-specific viewfinder images can be calculated, and the DOF corresponding to the viewfinder image with good focus can be selected as the representative DOF for each single area.
[0060] Figure 2a The division of specific individual regions of the photographed target using DOF is shown, and Figure 2b A graph is shown illustrating the selection of a representative DOF for a specific area using the focus of the viewfinder image of a single area at a particular DOF. In this case, since a representative DOF is selected after capturing the target object encompassing single areas at multiple DOFs, the capture process at each DOF can be termed a path. (Reference) Figure 2aIn a preferred embodiment, the individual regions can be divided into 7, and the DOFs can be divided into 5. For example, the first individual regions at the 5 DOFs, including the first DOF to the fifth DOF, can be represented as #1-1, #2-1, #3-1, #4-1, and #5-1, respectively, while the second individual regions can be represented as #1-2, #2-2, #3-2, #4-2, and #5-2. The third to seventh individual regions can also be represented in the same way at the 5 DOFs. Therefore, the capture process performed at the first DOF to the fifth DOF can be referred to as the first path to the fifth path, respectively.
[0061] The degree of focus for each individual area thus divided within the viewfinder can be calculated. Conventional viewfinders in this art calculate the DOF (Dispersion of Focus) of the captured image. For example, focus can be calculated by the magnitude of pixel intensity diffusion in the image captured by the viewfinder. If the viewfinder image is sharp and has good black-and-white distinction, the viewfinder pixel intensity diffusion appears large, indicating good focus.
[0062] refer to Figure 2b As a preferred embodiment, the actual height of each individual area in the target image can be obtained ( ) and representative DOFs selected for each individual region ( A curve graph (Z-axis).
[0063] When selecting a representative DOF for each individual region ( When ), it can be seen that when the representative DOF of each individual region ( When linked, the state from unfocused to representative DOF exists along the Z-axis ( (height movement)
[0064] In one aspect of the hyperspectral optical system of this disclosure, (c) the hyperspectral image capture unit can use a hyperspectral CCD (charge-coupled device) to capture a photographic target including a single region corresponding to a selected representative DOF.
[0065] Figure 3a This is a schematic diagram illustrating the repeated capture of a target, including individual regions, using a hyperspectral CCD based on a path representing the DOF. Figure 3b The image shown is a hyperspectral CCD image (scanned image) captured via the path.
[0066] In this scenario, capture using a hyperspectral CCD can be performed via line scan. For example, line scan capture can be performed along a path based on a representative DOF selected for each individual region.
[0067] In a hyperspectral optical system of one aspect of this disclosure, (d) the image correction unit can combine hyperspectral CCD images of individual regions corresponding to representative DOFs captured by a hyperspectral CCD to correct focus blur caused by non-uniform DOFs of the photographed target.
[0068] Figure 4a The process of combining the corrected images for each path is shown, and Figure 4b The corrected hyperspectral CCD image obtained by combining the corrected images of each path is shown.
[0069] Image correction software programs can be used to combine hyperspectral CCD images captured for each path to correct focus blur caused by non-uniform DOF of the target image. There are no particular limitations on the image correction software programs, as long as they operate using the logic of combining multiple images to correct focus blur.
[0070] refer to Figure 4a It can combine hyperspectral CCD images captured for each path with image correction software programs.
[0071] First, the hyperspectral CCD image captured by path 1 can be combined with images captured by paths 2 and 3, and then combined with images captured by paths 4 and 5 to create a complete image.
[0072] refer to Figure 4b By combining these images captured along the path, a final corrected hyperspectral CCD image can be obtained.
[0073] Figure 5 An image captured by a hyperspectral optical system employing autofocus technology, according to a preferred embodiment of one aspect of this disclosure, is shown. Figure 6 The image shown is captured by a conventional hyperspectral optical system (hyperspectral camera, Mitutoyo) without the application of autofocus technology.
[0074] In this case, there are no particular restrictions on the subject being photographed, but for example, if the DOF of the subject is uneven, it is better to confirm the effect of improving focus blur through autofocus.
[0075] It can be seen that, Figure 5 In the case of autofocus technology applied to the hyperspectral system, no focus blurring was observed, unlike in conventional hyperspectral CCDs that do not employ autofocus technology. Figure 6 In the process, blurring of focus was observed.
[0076] Autofocusing methods using hyperspectral optical systems
[0077] One aspect of this disclosure also relates to an autofocusing method using the hyperspectral optical system described above.
[0078] (A) A viewfinder imaging step, which divides the subject into multiple individual areas and then uses the viewfinder to capture a DOF-specific image of each individual area;
[0079] (B) A representative depth of field (DOF) derivation step for individual areas of the subject, which calculates the focus of each in the DOF-specific image of each individual area captured, and selects the DOF corresponding to the viewfinder image with good focus as the representative DOF of each individual area;
[0080] (C) A hyperspectral image acquisition step, which utilizes a hyperspectral CCD (charge-coupled device) to capture individual regions corresponding to selected representative DOFs; and
[0081] (D) Image correction step, which combines hyperspectral images of individual regions corresponding to representative DOF captured by hyperspectral CCD to correct focus blur caused by non-uniform DOF of the photographed target.
[0082] Although the present invention has been described with reference to limited embodiments and drawings, the invention is not limited thereto, and those skilled in the art can make various modifications and variations within the scope of the technical concepts of the invention and the equivalents of the following claims.
[0083] Explanation of reference numerals in the attached figures
[0084] 10: Viewfinder
[0085] 20: Shooting Target
[0086] 21: Individual Area
[0087] 22: Viewfinder Image
[0088] 22(n): Viewfinder image at the nth DOF
Claims
1. A hyperspectral optical system, comprising: (a) A viewfinder imaging unit that divides the subject into multiple separate regions and then uses the viewfinder to capture a depth-of-field (DOF) specific image of the subject including the separate regions; (b) A representative depth of field (DOF) derivation unit for a single area of the shooting target, wherein the representative depth of field (DOF) derivation unit calculates the focus of each of the captured DOF-specific viewfinder images and selects the DOF corresponding to the viewfinder image with good focus as the representative DOF of each single area; (c) A hyperspectral image capture unit that uses a hyperspectral CCD (charge-coupled device) to capture the individual region corresponding to a selected representative DOF; as well as (d) An image correction unit that combines hyperspectral images of the individual regions corresponding to the representative DOF captured by the hyperspectral CCD to correct focus blur caused by the non-uniform DOF of the photographed target.
2. The hyperspectral optical system according to claim 1, in, The target area was divided into 5 to 20 separate areas.
3. The hyperspectral optical system according to claim 1, in, Capture 3 to 10 DOF-specific images for each individual region.
4. The hyperspectral optical system according to claim 1, in, In the representative DOF export unit (b) for a separate area of the shooting target, the focus is calculated in the viewfinder.
5. An autofocusing method using a hyperspectral optical system, comprising: (A) A viewfinder imaging step, wherein the viewfinder imaging step divides the subject into multiple individual regions and then uses the viewfinder to capture depth-of-field (DOF) specific images of the individual regions; (B) A representative depth of field (DOF) derivation step for individual areas of the target being captured, wherein the representative depth of field (DOF) derivation step calculates the focus of each of the DOF-specific images of each individual area captured, and selects the DOF corresponding to the viewfinder image with good focus as the representative DOF of each individual area; (C) A hyperspectral image acquisition step, wherein the hyperspectral image acquisition step uses a hyperspectral CCD (charge-coupled device) to capture a single region corresponding to a selected representative DOF; and (D) Image correction step, which combines hyperspectral images of individual regions corresponding to the representative DOF captured by the hyperspectral CCD to correct focus blur caused by the non-uniform DOF of the photographed target.
6. The autofocusing method using a hyperspectral optical system according to claim 5, wherein, The target area is divided into 5 to 20 separate areas.
7. The autofocusing method using a hyperspectral optical system according to claim 5, wherein, Capture 3 to 10 DOF-specific images for each individual region.
8. The autofocusing method using a hyperspectral optical system according to claim 5, in, In step (B), the focus is calculated in the viewfinder.
Citation Information
Patent Citations
Double grating structured high efficiency transmission type diffraction grating which has broad wavelength bandwidth
KR1020240069400A
Apparatus for Depth-Resolved Hyperspectral Imaging
US20200200606A1