WO2014073288A1 - 画像処理装置、画像処理方法および画像処理プログラム - Google Patents
画像処理装置、画像処理方法および画像処理プログラム Download PDFInfo
- Publication number
- WO2014073288A1 WO2014073288A1 PCT/JP2013/076640 JP2013076640W WO2014073288A1 WO 2014073288 A1 WO2014073288 A1 WO 2014073288A1 JP 2013076640 W JP2013076640 W JP 2013076640W WO 2014073288 A1 WO2014073288 A1 WO 2014073288A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- amplitude component
- unit
- image processing
- rectangular
- Prior art date
Links
- 238000003672 processing method Methods 0.000 title claims description 6
- 238000006243 chemical reaction Methods 0.000 claims abstract description 34
- 238000000034 method Methods 0.000 claims description 76
- 230000008569 process Effects 0.000 claims description 44
- 238000007906 compression Methods 0.000 claims description 16
- 238000009795 derivation Methods 0.000 claims description 9
- 230000006835 compression Effects 0.000 claims description 8
- 230000009466 transformation Effects 0.000 abstract 3
- 238000000605 extraction Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 20
- 230000006870 function Effects 0.000 description 11
- 238000004422 calculation algorithm Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 230000015654 memory Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 125000002066 L-histidyl group Chemical group [H]N1C([H])=NC(C([H])([H])[C@](C(=O)[*])([H])N([H])[H])=C1[H] 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000003936 working memory Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/42—Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
- G06V10/431—Frequency domain transformation; Autocorrelation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/37—Determination of transform parameters for the alignment of images, i.e. image registration using transform domain methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
- G06V10/7515—Shifting the patterns to accommodate for positional errors
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20056—Discrete and fast Fourier transform, [DFT, FFT]
Definitions
- the present invention relates to an image processing apparatus, an image processing method, and an image processing program for estimating an image rotation amount.
- RIPOC is a technique that focuses on the characteristics of the amplitude component. That is, in RIPOC, an image is frequency-converted and its amplitude component is polar-coordinate converted to create a polar coordinate image having an angle ⁇ in the X direction and a radius r in the Y direction. Then, matching is performed between polar coordinate images.
- the deviation in the X direction corresponds to the angular deviation in the actual image. Therefore, the rotation amount can be estimated from the matching result.
- the shapes of the reference image (template) and the other images to be compared are both limited to squares. That is, in the conventional RIPOC, when a square image having the same size (number of pixels) in the X direction and the Y direction is targeted, and the image is rotated, the amplitude component when the frequency is converted is also rotated by the same amount. was the premise.
- the conventional RPOC does not consider the case where the shape of the image is not square, and has a problem that the shape of the template is limited to a square, for example. That is, there has been a problem that the rotation amount estimation process cannot be performed for images having different vertical and horizontal sizes.
- the present invention has been made in view of such problems, and provides an image processing apparatus, an image processing method, and an image processing program capable of estimating a rotation amount even when the vertical and horizontal sizes of an image are different. It is intended to provide.
- an image processing apparatus is an image processing apparatus that estimates a rotation amount between a first image and a second image, at least one of which is rectangular.
- the frequency conversion of the first image and the second image with respect to the rectangular space of the derivation unit for deriving the respective amplitude components and the first image and the second image is performed on the frequency space.
- An adjustment unit for performing an adjustment so that the short side size matches the long side size, and the amplitude component of the first image and the amplitude component of the second image adjusted by the adjustment unit The first image is calculated by calculating a position shift amount between the polar coordinate converted image of the amplitude component of the first image and the polar coordinate converted image of the amplitude component of the second image. To output the amount of rotation between the image and the second image And a matching unit.
- an image processing method is a method for estimating a rotation amount between a first image and a second image, at least one of which is rectangular, wherein the first image and the second image For the rectangular image of the first image and the second image, the size of the short side in the frequency space matches the size of the long side.
- a step of performing an adjustment to extend the image a step of performing a polar coordinate conversion between the amplitude component of the first image and the amplitude component of the second image after the adjustment, and a polar coordinate conversion image of the amplitude component of the first image And outputting a rotation amount between the first image and the second image by calculating a positional shift amount between the first image and the polar coordinate conversion image of the amplitude component of the second image.
- an image processing program is a program that causes a computer to execute a process of estimating a rotation amount between a first image and a second image, at least one of which is rectangular.
- the frequency conversion of the first image and the second image, the respective amplitude components being derived, and the rectangular size of the first image and the second image, the size of the short side in the frequency space Adjusting for expansion so that the length of the first image coincides with the size of the long side, converting the amplitude component of the first image and the amplitude component of the second image after the adjustment, and the first image
- a step of outputting a rotation amount between the first image and the second image by calculating a positional shift amount between the polar coordinate converted image of the amplitude component of the second image and the polar coordinate converted image of the amplitude component of the second image.
- the present invention it is possible to estimate the rotation amount between the first image and the second image even when the vertical and horizontal sizes of the images are different.
- FIGS. 7A and 7B are diagrams for explaining polar coordinate conversion, in which FIG.
- FIG. 7A shows an image before conversion
- FIG. 7B shows an image after conversion
- 3 is a block diagram illustrating a specific example of a functional configuration of the image processing apparatus according to the first embodiment.
- FIG. It is a flowchart showing the flow of corresponding point search operation
- FIG. 15 is a diagram for explaining the rotation when the image is rectangular.
- a rectangular area of 256 pixels (X direction) ⁇ 128 pixels (Y direction) has a stripe pattern of 1/64 (period / pixel) in real space in the X direction.
- the size serving as a base for calculating the spatial frequency is also different.
- the spatial frequency in the real space has been described, but there is a similar problem when the image is subjected to frequency conversion.
- the frequency distribution shape is distorted depending on the amount of rotation.
- FIG. 1 is a block diagram illustrating a specific example of the configuration of the image processing apparatus 100 according to the embodiment.
- FIG. 1 an example in which the image processing apparatus 100 according to the present embodiment is realized by a general personal computer is shown.
- an image processing apparatus 100 is mainly mounted on a computer having a general-purpose architecture.
- an image processing apparatus 100 includes, as main components, a CPU (Central Processing Unit) 102, a RAM (Random Access Memory) 104, a ROM (Read Only Memory) 106, and a network interface (I / F). ) 108, auxiliary storage device 110, display unit 120, input unit 122, and memory card interface (I / F) 124. Each component is communicably connected to each other via a bus 130.
- the CPU 102 controls the entire image processing apparatus 100 by executing various programs such as an operating system (OS) and an image processing program stored in the ROM 106, the auxiliary storage device 110, and the like.
- OS operating system
- image processing program stored in the ROM 106, the auxiliary storage device 110, and the like.
- the RAM 104 functions as a working memory for executing a program by the CPU 102, and temporarily stores various data necessary for executing the program.
- the ROM 106 stores an initial program (boot program) that is executed when the image processing apparatus 100 is started up.
- the network I / F 108 exchanges data with other devices (such as server devices) via various communication media. More specifically, the network I / F 108 is connected via a wired line such as Ethernet (registered trademark) (LAN (Local Area Network) or WAN (Wide Area Network)) and / or a wireless line such as a wireless LAN. Data communication.
- a wired line such as Ethernet (registered trademark) (LAN (Local Area Network) or WAN (Wide Area Network)) and / or a wireless line such as a wireless LAN.
- Ethernet registered trademark
- LAN Local Area Network
- WAN Wide Area Network
- the auxiliary storage device 110 typically includes a large-capacity magnetic storage medium such as a hard disk, and the like.
- the image processing program 112 for realizing various types according to the present embodiment, the search image 114 to be processed, the template image 300, and the like. Is stored. Further, the auxiliary storage device 110 may store a program such as an operating system.
- the search image 114 and the template image 300 are stored in order to search for a position on the search image 114 (second image) corresponding to the template image 300 (first image) registered in advance, for example.
- the main body of the image processing apparatus 100 does not have to have a function of capturing a subject.
- the image processing apparatus 100 acquires these images using a mechanism similar to a digital camera as will be described later, and uses these images in an arbitrary manner. You may make it input into. More specifically, these images are input to the image processing apparatus 100 via the network I / F 108 and the memory card I / F 124 described above.
- the display unit 120 displays a GUI (Graphical User Interface) screen provided by the operating system, an image generated by executing the image processing program 112, and the like.
- GUI Graphic User Interface
- the input unit 122 typically includes a keyboard, a mouse, a touch panel, and the like, and outputs the content of the instruction received from the user to the CPU 102 or the like.
- the memory card I / F 124 reads / writes data from / to various memory cards (nonvolatile storage media) 126 such as an SD (Secure Digital) card and a CF (Compact Flash (registered trademark)) card.
- various memory cards nonvolatile storage media
- SD Secure Digital
- CF Compact Flash
- the memory card I / F 124 is loaded with a memory card 126 storing an input image acquired by some device, and the input image read from the memory card 126 is stored (copied) in the auxiliary storage device 110.
- the image processing program 112 stored in the auxiliary storage device 110 is stored in a storage medium such as a CD-ROM (Compact Disk-Read Only Memory) and distributed, or distributed from a server device or the like via a network.
- the image processing program 112 calls a necessary module among program modules provided as part of an operating system executed by the image processing apparatus 100 (personal computer) at a predetermined timing and in order to realize the processing. May be.
- the image processing program 112 itself does not include a module provided by the operating system, and image processing is realized in cooperation with the operating system.
- the image processing program 112 may be provided by being incorporated in a part of some program instead of a single program.
- the image processing program 112 itself does not include a module that is commonly used in the program, and image processing is realized in cooperation with the program. Even such an image processing program 112 that does not include some modules does not depart from the spirit of the image processing apparatus 100 according to the present embodiment.
- image processing program 112 may be realized by dedicated hardware.
- the image processing apparatus 100 is not limited to that realized by a general personal computer as illustrated in FIG. 1, and other configurations similar to a digital camera or a terminal such as a mobile phone It may be realized by a device or the like. Further, it may be in the form of a so-called cloud service in which at least one server device realizes processing according to the present embodiment.
- the user transmits the search image 114 and the template image 300 to the server device (cloud side) using his / her terminal (such as a personal computer or a smartphone), and the transmitted search image 114 and template image 300 are transmitted.
- the server device side performs image processing according to the present embodiment is assumed.
- it is not necessary for the server device side to perform all functions (processing), and the user side terminal and the server device may cooperate to realize the image processing according to the present embodiment.
- a corresponding point search operation involving rotation amount estimation is performed. That is, in the corresponding point search operation, the image processing apparatus 100 detects how much another image (hereinafter referred to as a search image) to be compared is rotated with respect to a reference image (hereinafter referred to as a template image). The rotation amount estimation process is executed.
- the angle estimation process In the rotation amount estimation process, generally, when using a template matching method such as SAD (Sum of Absolute Differences) or SSD (Sum of Squared Differences), the angle is estimated by rotating and matching the template image. To do. However, in this method, if the angle is estimated with an accuracy of 1 °, for example, the template has to be rotated 360 times each of 0 ° to 359 ° and matching is performed, so that the processing time becomes long. For this reason, the image processing apparatus 100 according to the present embodiment employs an angle estimation method based on the RIPOC (Rotation Invariant Phase Only Correlation) method to estimate the rotation amount and perform a corresponding point search operation. . In the RIPOC method, the amount of rotation between a plurality of images is estimated by frequency-converting an image and performing polar coordinate conversion of the amplitude component and collating them.
- RIPOC Ratation Invariant Phase Only Correlation
- FIG. 2 is a schematic diagram showing a rotation amount estimation processing algorithm according to the first embodiment.
- the algorithm of the rotation amount estimation process according to the first embodiment includes amplitude component derivation processes 202 and 212, adjustment processes 203 and 213, compression processes 204 and 214, polar coordinate conversion process 206, 216 and a matching process 208.
- the amplitude component derivation processes 202 and 212 obtain the amplitude component by converting the search image and the template image into frequency components (amplitude component and phase component), respectively.
- frequency components amplitude component and phase component
- Fourier transform is used, but Laplace transform or the like may be used.
- the phase component is not necessarily required, and thus may not be calculated.
- the search image and the template image are rectangles, and at least one of them is a rectangle having different vertical and horizontal sizes.
- the adjustment processes 203 and 213 are processes for aligning the one with fewer frequency components (hereinafter referred to as the short side) in the frequency space to the one with more frequency components (hereinafter referred to as the long side). To do. That is, processing for matching the vertical and horizontal sizes in the frequency space is performed. This processing is equivalent to matching the unit frequency, that is, the size (number of pixels) at which the frequency is 1 in the vertical and horizontal directions of the same image without changing the period per pixel in real space. Since the period per pixel is not changed, information on the frequency component (amplitude component) is maintained before and after the adjustment process.
- 3 and 4 are diagrams showing specific examples of amplitude component images before and after adjustment according to the first embodiment.
- the vertical axis and the horizontal axis indicate frequency components (amplitude components) in the XY direction. As shown in FIG. 3 and FIG. 4, these are matched by extending the smaller frequency component (short side) to the same size (long side) and the same size (long side).
- the interpolation method for the portion without information at the time of expansion is performed using the average value of the peripheral coordinate values, but other methods may be used. By this adjustment, the frequency component is maintained even when the rotation is performed, and the angle can be estimated.
- the adjustment processes 203 and 213 may be executed on at least one of the template image and the search image. That is, if the template image is a rectangle having different vertical and horizontal sizes and the search image is a square, adjustment processing only needs to be performed for the template image.
- 5 and 6 are diagrams for explaining the influence of rotation in matching between the case where the search image is rectangular and the case where the search image is square.
- the image processing apparatus 100 rotates the template image and then performs a matching process with the search image to thereby position the template image on the search image. Is estimated.
- a case where a rectangular image (region indicated by a frame in the drawing) including both eyes of a person is set as a template image will be described. If the search image is a rectangle of the same size, and the image is not rotated, it is possible to set the same area as the template image in the right diagram of FIG. 5 and perform matching with the template image. Can do. On the other hand, when the image is rotated, the area including both eyes cannot be set as the search image as shown in the left diagram of FIG. Matching with will not go well.
- the search image when the search image is a square, it is possible to suppress the possibility that the subject is completely closed as shown in FIG. That is, not only when the image is not rotated (right diagram in FIG. 6) but also when the image is rotated (right diagram in FIG. 6), a region including both eyes can be set as the search image. it can. That is, in this case, the template image is rectangular and the search image is square. In this way, even if the template image and the search image have different shapes, the rotation amount can be estimated by adjusting the template image using the adjustment processes 203 and 213. For this reason, for example, by using a square search image for a rectangular template image, it is possible to perform a search because the object is not seen when the rotation amount is large while suppressing the use of unnecessary objects in the template image. You can avoid the situation of disappearing.
- Compression processing 204, 214 compresses the amplitude component of the search image and the template image, respectively.
- Examples of the compression method include a logarithmization method, a square root calculation method, and a N-th power with a predetermined value N less than 1.
- Polar coordinate conversion processes 206 and 216 convert the compressed amplitude components of the search image and the template image into polar coordinate components, respectively. By this conversion, the rotation angle is expressed as a coordinate point on two-dimensional coordinates.
- FIGS. 7A and 7B are diagrams for explaining polar coordinate conversion, in which FIG. 7A shows an image before conversion, and FIG. 7B shows an image after conversion.
- the polar component conversion processing 206 and 216 converts the amplitude component of the image into a polar coordinate component represented by an angle theta in the X direction and a radius r in the Y direction.
- FIG. 7 shows polar coordinate conversion of an image in real space for easy understanding.
- the matching processing 208 associates the results of the polar coordinate conversion output from the polar coordinate conversion processing 206 and 216, respectively, and obtains a positional deviation amount.
- one axis represents the angle theta and the other axis represents the radius r, so that the shift amount on the theta side represents the rotation amount.
- the matching processing 208 detects a deviation amount by calculating a POC value (similarity) distribution having the same size as the image size and specifying a peak position of the distribution in one processing. . That is, the matching processing 208 specifies the position where the similarity is the highest among the results of the polar coordinate conversion, and outputs the rotation amount corresponding to the estimated rotation angle.
- POC value similarity
- the adjustment processes 203 and 213 are performed before the compression processes 204 and 214.
- the adjustment can be performed based on finer amplitude components, and the accuracy can be improved.
- these processing orders are not limited to the order illustrated in FIG. That is, the adjustment processes 203 and 213 may be performed after the compression processes 204 and 214, or may be performed after the polar coordinate conversion processes 206 and 216. Since the data before the compression processes 204 and 214 are easily affected by illumination changes and shading, the adjustment processes 203 and 213 are performed after the compression processes 204 and 214, so that fluctuation due to such noise can be suppressed. Therefore, preferably, the processing order is switched according to the subject and the environment.
- FIG. 8 is a block diagram showing a specific example of a functional configuration of the image processing apparatus 100 according to the first embodiment for performing the corresponding point search operation with the rotation amount estimation.
- Each function in FIG. 8 is a function mainly formed on the CPU 102 when the CPU 102 of the image processing apparatus 100 reads out the program stored in the ROM 106 or the auxiliary storage device 110 to the RAM 104 and executes the program.
- At least a part may be realized by the hardware configuration shown in FIG.
- the auxiliary storage device 110 is provided with an image storage unit 111 that is a storage area for storing the search image 114 and the template image 300.
- image processing apparatus 100 includes an angle estimation unit 10, an image reading unit 11, a position estimation unit 17, and an output unit 18 as its main functional configuration.
- the image reading unit 11 reads the search image 114 and the template image 300 from the image storage unit 111 and inputs them to the angle estimation unit 10.
- the angle estimation unit 10 includes an amplitude component deriving unit 12 for performing amplitude component deriving processes 202 and 212, an adjusting unit 13 for performing adjustment processes 203 and 213, and a compressing unit 14 for performing compression processes 204 and 214. And a conversion unit 15 for performing polar coordinate conversion processing 206 and 216 and a matching unit 16 for performing matching processing 208.
- the position estimation unit 17 specifies the rotation angle between the search image 114 and the template image 300 based on the result of the matching process in the matching unit 16 and corrects the angle so that the template image 300 is rotated by the angle.
- the position of the search image on the template image 300 is estimated, and the position information is input to the output unit 18.
- the output unit 18 may output the position information by, for example, displaying the position information on the display unit 120, or may output the position information to another device from the network I / F 108 via various communication media.
- FIG. 9 is a flowchart showing the flow of the corresponding point search operation in the image processing apparatus 100 according to the first embodiment. The operation shown in the flowchart of FIG. 9 is performed by causing the CPU 102 of the image processing apparatus 100 to read out and execute a program stored in the ROM 106, the auxiliary storage device 110, and the like to the RAM 104, and exhibit each function of FIG. Realized.
- a template image and a search image are acquired (steps S101 and S103). At least one of the template image and the search image is a rectangular image.
- each image is frequency-converted to derive its amplitude component (step S105), and at least one image (rectangular image) is adjusted by adjusting the vertical and horizontal sizes in the frequency space (step S105). S107).
- step S109 the amplitude components of the search image and the template image are compressed (step S109), and the compressed amplitude components of the search image and the template image are converted into polar coordinate components (step S111). By matching these polar coordinate images, the displacement amount of these image positions is calculated (step S113).
- the rotation amount (rotation angle) between the template image and the search image is specified from the shift amount obtained in step S113 (step S115).
- one image for example, a template image
- the rotation angle for example, a template image
- the search image for example, a template image
- Position information representing the position of the template image on the search image estimated by the matching process is output as information representing the corresponding point (step S121).
- FIG. 10 is a schematic diagram showing an algorithm of the rotation amount estimation process according to the second embodiment.
- the algorithm of the rotation amount estimation process according to the second embodiment includes adjustment processes 201 and 211, amplitude component derivation processes 202 and 212, compression processes 204 and 214, polar coordinate conversion process 206, 216 and a matching process 208.
- the adjustment processing 201 and 211 when the vertical and horizontal sizes of the images are different in real space, the adjustment processing 201 and 211 has a unit frequency, that is, a frequency, without changing the period per pixel. Match the size (number of pixels) of 1 vertically and horizontally. That is, the processing for adding image information is performed so that the size of the short side matches the size of the long side without changing the cycle per pixel in the real space. Then, the amplitude component derivation processes 202 and 212 frequency-convert the adjusted image and derive the amplitude component.
- FIG. 11 is a diagram illustrating a specific example of a real space image before and after adjustment according to the second embodiment.
- the side having the smaller number of pixels hereinafter referred to as the short side
- the long side is made the same size as the one having the larger number (hereinafter referred to as the long side).
- a method is mentioned.
- the image is regarded as a periodic image in which the image end is connected to the opposite end, and the size is enlarged by adding the image at the opposite end to the image end.
- the amplitude component after frequency conversion is expanded so that the size of the short side matches the size of the long side in the frequency space, as in FIG. .
- extension method in the adjustment process is not limited to the method illustrated in FIG. 11, and may be another method.
- 12 and 13 are diagrams showing specific examples of real space images before and after adjustment by another method.
- FIG. 12 there is a method of enlarging the size by folding the image (adding the inverted image).
- FIG. 13 there is a method of expanding the size by filling with a predetermined value (for example, 0) (adding a predetermined value (for example, 0)).
- the signal component is calculated as a periodic function. For example, since an image whose edge portion is a predetermined value (for example, 0) as shown in FIG. 13B has a signal that changes rapidly, a frequency component that does not originally exist is observed during frequency conversion. Problems can occur. However, since the influence can be reduced by applying a window function such as a Hanning window before Fourier transform, as an enlargement method in the adjustment process, enlargement by adding a predetermined value as shown in FIG. The method can be adopted.
- a window function such as a Hanning window before Fourier transform
- the functional configuration of the image processing apparatus 100 according to the second embodiment is substantially the same as the functional configuration of the image processing apparatus 100 according to the first embodiment shown in FIG.
- the adjustment unit 13 performs adjustment processing 201 and 211, and inputs the result to the amplitude component deriving unit 12.
- the amplitude component calculated by the amplitude component deriving unit 12 is input to the compression unit 14.
- FIG. 14 is a flowchart showing the flow of the corresponding point search operation in the image processing apparatus 100 according to the second embodiment. 14 also causes the CPU 102 of the image processing apparatus 100 to read out the program stored in the ROM 106, the auxiliary storage device 110, and the like to the RAM 104 and execute the program, thereby exercising the functions shown in FIG. It is realized by.
- step S104 when the processing of steps S101 to S103 is performed and the template image and the search image are acquired, at least one image (rectangular shape) is obtained.
- the above-described adjustment is performed on the real image in the real space (step S104).
- the frequency of the adjusted image is converted, and the amplitude component is derived (step S105).
- step S105 Thereafter, the same processing as that of the image processing apparatus 100 according to the first embodiment after S109 is performed, so that the position information is output.
- this adjustment may be performed in the frequency space or in the real space.
- the amount of calculation can be reduced as compared with the case of performing in the real space.
- the interpolation can be performed with higher accuracy than the interpolation at the time of adjustment in the frequency space, the accuracy can be improved.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
Description
上記問題を解決する方法の一例として、特開平10-124667号公報(以下、特許文献1)に開示されているような、RIPOC(Rotation Invariant Phase Only Correlation:回転不変位相限定相関)などと呼ばれる、画像を周波数変換し、その振幅成分を極座標変換して回転量を推定する手法が提案されている。
まず、本発明の実施の形態の説明の前に、画像が長方形である場合の回転について説明する。
図1は、実施の形態にかかる画像処理装置100の構成の具体例を示すブロック図である。図1の例では、本実施の形態にかかる画像処理装置100を一般的なパーソナルコンピューターにより実現した場合の例が示されている。
本実施の形態にかかる画像処理装置100では、回転量推定を伴う対応点探索動作を行なう。すなわち、画像処理装置100では、対応点探索動作において、基準となる画像(以下、テンプレート画像)に対し、比較対象となる他の画像(以下、探索画像)がどの程度回転しているかを検出する、回転量推定処理を実行する。
図2は、第1の実施の形態に従う回転量推定処理のアルゴリズムを示す模式図である。図2を参照して、第1の実施の形態に従う回転量推定処理のアルゴリズムは、振幅成分導出処理202,212と、調整処理203,213と、圧縮処理204,214と、極座標変換処理206,216と、マッチング処理208とを含む。
図8は、上記回転量推定を伴う対応点探索動作を行なうための、第1の実施の形態にかかる画像処理装置100の機能構成の具体例を示すブロック図である。図8の各機能は、画像処理装置100のCPU102がROM106や補助記憶装置110などに記憶されているプログラムをRAM104に読み出して実行することで、主に、CPU102上に形成される機能であるが、少なくとも一部が、図1に示されたハードウェア構成によって実現されてもよい。
図9は、第1の実施の形態にかかる画像処理装置100での対応点探索動作の流れを表わすフローチャートである。図9のフローチャートに表わされた動作は、画像処理装置100のCPU102がROM106や補助記憶装置110などに記憶されているプログラムをRAM104に読み出して実行し、図8の各機能を発揮させることによって実現される。
図10は、第2の実施の形態に従う回転量推定処理のアルゴリズムを示す模式図である。図10を参照して、第2の実施の形態に従う回転量推定処理のアルゴリズムは、調整処理201,211と、振幅成分導出処理202,212と、圧縮処理204,214と、極座標変換処理206,216と、マッチング処理208とを含む。
第2の実施の形態にかかる画像処理装置100の機能構成は、図8に示された第1の実施の形態にかかる画像処理装置100の機能構成と概ね同じものである。第2の実施の形態にかかる画像処理装置100では、角度推定部10において、調整部13は調整処理201,211を行ない、その結果を振幅成分導出部12に入力する。振幅成分導出部12で算出された振幅成分は、圧縮部14に入力される。
図14は、第2の実施の形態にかかる画像処理装置100での対応点探索動作の流れを表わすフローチャートである。図14のフローチャートに表わされた動作もまた、画像処理装置100のCPU102がROM106や補助記憶装置110などに記憶されているプログラムをRAM104に読み出して実行し、図8の各機能を発揮させることによって実現される。
本実施の形態にかかる画像処理装置100では、対応点探索動作の際の回転量推定において、探索画像およびテンプレート画像のうちの少なくとも一方の画像として縦横のサイズが異なる長方形の画像を用い、その画像について、空間周波数の解像度、すなわち空間周波数を算出する際のベースとなるサイズを縦辺と横辺とで同じ値となるよう調整する。これにより、画像を周波数変換した際の振幅成分を用いた回転量推定においてテンプレート画像または探索画像の形状に関わらずにこれら画像間での角度推定が可能となる。
Claims (8)
- 少なくとも一方が長方形である第1の画像と第2の画像との間の回転量を推定する画像処理装置であって、
前記第1の画像と前記第2の画像とを周波数変換し、それぞれの振幅成分を導出するための導出部と、
前記第1の画像と前記第2の画像とのうちの長方形の画像について、周波数空間上における短辺のサイズが長辺のサイズと一致するように伸長するための調整を行なうための調整部と、
前記調整部によって調整された、前記第1の画像の振幅成分と前記第2の画像の振幅成分とを極座標変換するための変換部と、
前記第1の画像の振幅成分の極座標変換画像と前記第2の画像の振幅成分の極座標変換画像との位置のずれ量を算出することで、前記第1の画像と前記第2の画像との間の回転量を出力するためのマッチング部とを備える、画像処理装置。 - 前記調整部は、前記長方形の画像が前記導出部によって周波数変換された後に、周波数空間上で、短辺を長辺と一致するように補間することによって伸長する、請求項1に記載の画像処理装置。
- 前記調整部は、前記長方形の画像が前記導出部によって周波数変換される前に、実空間上で、1画素当たりの周期を変更することなく、短辺のサイズが長辺のサイズに一致するように、画像情報を追加する、請求項1に記載の画像処理装置。
- 前記第1の画像の振幅成分と前記第2の画像の振幅成分とを圧縮するための圧縮部をさらに備える、請求項1~3のいずれか1項に記載の画像処理装置。
- 前記圧縮部は、前記調整部によって調整された後に、前記第1の画像の振幅成分と前記第2の画像の振幅成分とを圧縮する、請求項4に記載の画像処理装置。
- 前記第1の画像の振幅成分と前記第2の画像の振幅成分とを圧縮するための圧縮部をさらに備え、
前記調整部は、前記長方形の画像が圧縮された後に、周波数空間上で前記調整を行なう、請求項2に記載の画像処理装置。 - 少なくとも一方が長方形である第1の画像と第2の画像との間の回転量を推定する方法であって、
前記第1の画像と前記第2の画像とを周波数変換し、それぞれの振幅成分を導出するステップと、
前記第1の画像と前記第2の画像とのうちの長方形の画像について、周波数空間上における短辺のサイズが長辺のサイズと一致するように伸長するための調整を行なうステップと、
前記調整後に、前記第1の画像の振幅成分と前記第2の画像の振幅成分とを極座標変換するステップと、
前記第1の画像の振幅成分の極座標変換画像と前記第2の画像の振幅成分の極座標変換画像との位置のずれ量を算出することで、前記第1の画像と前記第2の画像との間の回転量を出力するステップとを備える、画像処理方法。 - 少なくとも一方が長方形である第1の画像と第2の画像との間の回転量を推定する処理をコンピューターに実行させるプログラムであって、
前記第1の画像と前記第2の画像とを周波数変換し、それぞれの振幅成分を導出するステップと、
前記第1の画像と前記第2の画像とのうちの長方形の画像について、周波数空間上における短辺のサイズが長辺のサイズと一致するように伸長するための調整を行なうステップと、
前記調整後に、前記第1の画像の振幅成分と前記第2の画像の振幅成分とを極座標変換するステップと、
前記第1の画像の振幅成分の極座標変換画像と前記第2の画像の振幅成分の極座標変換画像との位置のずれ量を算出することで、前記第1の画像と前記第2の画像との間の回転量を出力するステップとを前記コンピューターに実行させる、画像処理プログラム。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/437,621 US20150262028A1 (en) | 2012-11-07 | 2013-10-01 | Image processing apparatus, image processing method and image processing program |
JP2013555693A JP5500328B1 (ja) | 2012-11-07 | 2013-10-01 | 画像処理装置、画像処理方法および画像処理プログラム |
EP13852732.0A EP2919196A4 (en) | 2012-11-07 | 2013-10-01 | IMAGE PROCESSING DEVICE, METHOD, AND PROGRAM |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2012245222 | 2012-11-07 | ||
JP2012-245222 | 2012-11-07 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014073288A1 true WO2014073288A1 (ja) | 2014-05-15 |
Family
ID=50684405
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2013/076640 WO2014073288A1 (ja) | 2012-11-07 | 2013-10-01 | 画像処理装置、画像処理方法および画像処理プログラム |
Country Status (4)
Country | Link |
---|---|
US (1) | US20150262028A1 (ja) |
EP (1) | EP2919196A4 (ja) |
JP (1) | JP5500328B1 (ja) |
WO (1) | WO2014073288A1 (ja) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108133492B (zh) * | 2016-12-01 | 2022-04-26 | 京东方科技集团股份有限公司 | 图像匹配方法、装置和系统 |
CN113496476A (zh) * | 2020-04-02 | 2021-10-12 | 北京东舟技术股份有限公司 | 花屏图像判断方法、设备及计算机存储介质 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10124667A (ja) | 1996-08-26 | 1998-05-15 | Yamatake Honeywell Co Ltd | パターン照合装置 |
JP2002063581A (ja) * | 2000-08-23 | 2002-02-28 | Nippon Telegr & Teleph Corp <Ntt> | 画像の回転角度検出方法および回転角度補正方法 |
JP2004192633A (ja) * | 2002-11-25 | 2004-07-08 | Yamatake Corp | パターン照合装置 |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8942512B2 (en) * | 2011-12-24 | 2015-01-27 | Ecole De Technologie Superieure | Methods and systems for processing a first image with reference to a second image |
-
2013
- 2013-10-01 WO PCT/JP2013/076640 patent/WO2014073288A1/ja active Application Filing
- 2013-10-01 US US14/437,621 patent/US20150262028A1/en not_active Abandoned
- 2013-10-01 EP EP13852732.0A patent/EP2919196A4/en not_active Withdrawn
- 2013-10-01 JP JP2013555693A patent/JP5500328B1/ja not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10124667A (ja) | 1996-08-26 | 1998-05-15 | Yamatake Honeywell Co Ltd | パターン照合装置 |
JP2002063581A (ja) * | 2000-08-23 | 2002-02-28 | Nippon Telegr & Teleph Corp <Ntt> | 画像の回転角度検出方法および回転角度補正方法 |
JP2004192633A (ja) * | 2002-11-25 | 2004-07-08 | Yamatake Corp | パターン照合装置 |
Non-Patent Citations (1)
Title |
---|
See also references of EP2919196A4 |
Also Published As
Publication number | Publication date |
---|---|
JP5500328B1 (ja) | 2014-05-21 |
US20150262028A1 (en) | 2015-09-17 |
EP2919196A1 (en) | 2015-09-16 |
EP2919196A4 (en) | 2016-06-22 |
JPWO2014073288A1 (ja) | 2016-09-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11164323B2 (en) | Method for obtaining image tracking points and device and storage medium thereof | |
JP6784704B2 (ja) | 画像処理方法及び装置 | |
US10123024B2 (en) | Image processing methods and image processing apparatuses | |
JP5294343B2 (ja) | 画像位置合わせ処理装置、領域拡張処理装置及び画質改善処理装置 | |
KR102268878B1 (ko) | 카메라 교정 | |
CN106560840B (zh) | 一种图像信息识别处理方法及装置 | |
JP6512100B2 (ja) | 画像処理を実行する情報処理装置及び画像処理方法 | |
US9613404B2 (en) | Image processing method, image processing apparatus and electronic device | |
US20180247430A1 (en) | Display control method and display control apparatus | |
JP5500328B1 (ja) | 画像処理装置、画像処理方法および画像処理プログラム | |
JP6102938B2 (ja) | 画像処理装置、及び、画像処理方法 | |
US20150345936A1 (en) | Image Processing Apparatus, Image Processing Method And Image Processing Program | |
WO2013089261A1 (ja) | 画像処理システム及び画像処理方法 | |
JP2017049686A (ja) | 画像処理装置 | |
CN113763233B (zh) | 一种图像处理方法、服务器及拍照设备 | |
JP2019220032A (ja) | 原画像を目標画像に基づいて変形した表示画像を生成するプログラム、装置及び方法 | |
CN113167568B (zh) | 坐标计算装置、坐标计算方法和计算机可读记录介质 | |
JP6056354B2 (ja) | 画像処理装置、画像処理方法および画像処理プログラム | |
JP2018010359A (ja) | 情報処理装置、情報処理方法、およびプログラム | |
CN106296800B (zh) | 信息处理方法及电子设备 | |
JP2014029677A (ja) | 画像処理装置、画像処理方法および画像処理プログラム | |
TWI779922B (zh) | 擴增實境處理裝置以及方法 | |
CN115564644B (zh) | 图像数据的处理方法、相关设备以及计算机存储介质 | |
CN107808411B (zh) | 多视角运动法三维重建方法 | |
JP2012169735A (ja) | 画像処理装置、画像処理方法及びプログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ENP | Entry into the national phase |
Ref document number: 2013555693 Country of ref document: JP Kind code of ref document: A |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13852732 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14437621 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2013852732 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |