JP2017016511A - Distortion correction image processor and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device and a program which efficiently compress an LUT for distortion correction to reduce information quantity.SOLUTION: An image processor comprises image correction means 10 for correcting distortion of image data. The image correction means has a correction value storage part 11, an LUT expansion part 12, and a coordinate conversion part 13. The correction value storage part stores an offset code table 11a and a compression LUT 11b. The offset code table associates an offset value between a coordinate value before correcting distortion and a coordinate value after correcting distortion with a predetermined offset code. The compression LUT associates the offset code with a coordinate value which is compressed referring to an offset value corresponding to the offset code. The LUT expansion part generates an LUT 11c for correction in which the coordinate value before correcting distortion is associated with the coordinate value after correcting distortion based on the offset code table and the compression LUT. The coordinate conversion part corrects a coordinate value of image data based on the LUT for correction.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an image processing apparatus and a program having a function for correcting distortion of image data. More specifically, in the present invention, an LUT (Look Up Table) for correcting distortion in units of pixels is created in advance, and this LUT is referred to when image data is input. The present invention relates to an apparatus and a program for performing image distortion correction processing.

  Since there is aberration in a general optical lens, it is known that distortion corresponding to the optical shape of the optical lens occurs in image data (photographs and moving images) acquired by the optical lens. As the distortion according to the optical shape, a pincushion shape or barrel shape distortion is known. In addition, there are individual differences in optical lenses, and the shape of distortion that occurs in image data may differ from one optical lens to another. For this reason, when displaying the image data acquired by the optical lens on a flat display, it is necessary to perform processing for returning the image data having distortion to the original flat shape. On the other hand, undistorted image data (computer graphics) generated by a GPU (Graphics Processing Unit) is displayed on a distorted display, for example, a transparent head-up display used in an in-car car navigation system. In some cases. In this case, the image data generated by the GPU is subjected to correction processing that gives distortion corresponding to the display so that the user does not feel the distortion even if the display has distortion. An image can be displayed. As described above, correction processing for distorting image data having no distortion may be performed.

  Here, conventionally, it is known to use an LUT in processing for correcting distortion of image data (Patent Document 1). The distortion correction LUT stores, for each pixel, a coordinate value of data before distortion correction and a coordinate value after distortion correction in association with each other. Therefore, by referring to this LUT when processing pixels in a region where distortion occurs, individual differences in lens distortion occurring during mass production can be corrected in units of one pixel. In this way, by setting the LUT, it is possible to correct the image distortion according to the optical shape of the optical lens or the like, and obtain a coordinate value in which the distortion is eliminated.

JP 2013-179513 A

  However, when performing distortion correction using an LUT, it is necessary to define a corrected coordinate value for each pixel. Therefore, the distortion correction LUT has a problem in that the amount of information (data amount) is large and the storage space of the memory is compressed. In addition, if the amount of information in the LUT is large, it takes time to read the LUT, which may cause processing delay and performance degradation in the entire image processing apparatus. Accordingly, an object of the present invention is to reduce the amount of information by efficiently compressing a distortion correction LUT to improve the LUT read performance, the processing speed and the processing performance of the entire image processing apparatus.

  As a result of intensive studies on means for solving the problems of the conventional invention, the inventors of the present invention associate an offset value (difference value) of coordinate values before and after distortion correction with a predetermined offset code, and this offset The inventor has obtained the knowledge that a LUT for distortion correction is compressed using a code. By compressing the LUT in this way, the amount of information in the LUT can be efficiently reduced, so that the processing speed and processing performance of the image processing apparatus can be improved. Further, by compressing the LUT, processing information for each other pixel (for example, mask information related to clipping, image ID, information for brightness adjustment) can be stored in the surplus space generated by the compression. It becomes like this. Then, the present inventors have conceived that the problems of the conventional invention can be solved based on the above knowledge, and have completed the present invention. More specifically, the present invention has the following configuration.

The first aspect of the present invention relates to an image processing apparatus.
The image processing apparatus according to the present invention includes image correction means 10 for correcting distortion of image data.
The image correction unit 10 includes a correction value storage unit 11, an LUT development unit 12, and a coordinate conversion unit 13.
The correction value storage unit 11 stores an offset code table 11a and a compression LUT 11b. In the offset code table 11a, the offset value between the coordinate value before distortion correction and the coordinate value after distortion correction is associated with a predetermined offset code. The compression LUT 11b associates the offset code with the coordinate value compressed by referring to the offset value corresponding to the offset code. That is, by using the offset code, the compressed coordinate value in the compression LUT 11b is reduced in the amount of information compared to the coordinate value after normal distortion correction.
Based on the offset code table 11a and the compressed LUT 11b, the LUT development unit 12 generates a correction LUT 11c that associates the coordinate values before distortion correction and the coordinate values after distortion correction. That is, the correction LUT 11c is compressed based on the offset code to become a compression LUT 11b. Therefore, when necessary, the LUT decompressing unit 12 decompresses the compressed LUT 11b based on the offset code to generate a correction LUT 11c. In this sense, it can be said that the compression LUT 11b is reversibly compressed.
The coordinate conversion unit 13 corrects the coordinate value of the image data based on the correction LUT 11c.

  As described above, the storage space in the memory can be effectively used by compressing the correction LUT using the offset code and storing it in the memory (correction value storage unit 11). It is also possible to store other pixel unit processing information in the surplus space generated by compressing the LUT. Further, by reducing the amount of information in the LUT, it is possible to improve the read performance of the LUT, and consequently improve the processing speed and processing performance of the entire image processing apparatus.

  In the image processing apparatus of the present invention, the offset code table 11a associates the offset values of the integer part of the coordinate value before distortion correction and the integer part of the coordinate value after distortion correction with a predetermined offset code. May be. Further, the compression LUT 11b may be one in which the offset code is associated with the decimal part of the coordinate value obtained by subtracting the offset value corresponding to the offset code. As a result, the integer part of the image coordinate value is indicated by the offset code, and only the decimal part is held in the compression LUT 11b. In this way, the offset code can be defined with a small amount of information by dividing the decimal part into an integer part and a decimal part and changing the encoding method (compression method). Further, the correction LUT 11c is easily generated by decompressing the compression LUT 11b.

  In the image processing apparatus of the present invention, in addition to the compressed coordinate value and offset code, the compression LUT 11b is associated with an extended code that indicates processing information used for image processing at the coordinate value. preferable. According to the present invention, since the compression LUT 11b can be constructed with a small amount of information, other processing information can be stored in the memory space in which the amount of information is compressed.

  In the image processing apparatus of the present invention, when the image data after distortion correction is vertically symmetrical or laterally symmetric, the correction LUT 11c holds only one of the upper and lower or left and right coordinate values with the symmetry axis as a boundary. It is good. The amount of information stored in the memory can be reduced by half by holding one of the upper and lower or left and right coordinate values and deriving the other coordinate value by a simple calculation.

  In the present invention, the coordinate conversion unit 13 converts the coordinate value after distortion correction held in the correction LUT 11c and the coordinate value obtained by applying a predetermined distortion correction formula to the coordinate value of the image data. Based on this, the coordinate value of the image data may be corrected. In this way, by correcting the coordinate value using both the LUT value and the distortion correction formula, even a complicated calculation can be processed at high speed. In addition, the accuracy of distortion correction can be increased.

  The second aspect of the present invention relates to a computer program for correcting distortion of image data. The program according to the present invention causes some functions of a general-purpose computer to function as the image processing apparatus according to the first aspect described above.

  According to the present invention, the distortion correction LUT can be efficiently compressed, and the amount of information can be reduced. In addition, by reducing the amount of information in the LUT, it is possible to improve the LUT read performance by the image processing apparatus and the processing speed and processing performance of the entire image processing apparatus.

FIG. 1 is a block diagram showing a configuration example of an image processing apparatus according to the present invention. FIG. 2 shows an example of the correction LUT. FIG. 3 shows an example of the offset code table. FIG. 4 shows an example of a compressed LUT and an example of a compressed bit configuration. FIG. 5 is a flowchart showing an example of processing by the control unit. FIG. 6 shows a configuration example of compressed bits including an extension code. FIG. 7 shows a compression example of the correction LUT when the image data is symmetrical.

  Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. This invention is not limited to the form demonstrated below, The thing suitably changed in the range obvious to those skilled in the art from the following forms is also included.

  The present invention relates to an image processing apparatus and a program having a function for correcting distortion of image data. In the present specification, the processing for correcting the distortion of the image data is performed in accordance with the processing for eliminating the distortion of the image data having distortion, in accordance with the flat display without distortion, and in accordance with the distorted display. Both processes for distorting image data having no distortion are included. For this reason, it can be said that the processing of “distortion correction” and “correction” is image processing for adapting the display image and the shape of the display.

  FIG. 1 is a block diagram showing an outline of an image processing apparatus 100 according to the present invention. The image processing apparatus 100 of the present invention has a function for correcting distortion of image data. As shown in FIG. 1, the image processing apparatus 100 includes an image correction unit 10, an image data input unit 20, a line buffer 30, a filter unit 40, an image output unit 50, and a control unit 60. The control unit 60 performs overall control of each element constituting the image processing apparatus 100. The line buffer 30 can be replaced with an external memory provided outside the image processing apparatus 100.

  As shown in FIG. 1, image data is input to the image processing apparatus 100 from an input device 200 such as a camera or an external memory. This image data includes image distortion. For example, since there is an aberration in a general optical lens, distortion corresponding to the optical shape of the optical lens is generated in image data (photographs and moving images) input from the camera. As the distortion according to the optical shape, a pincushion shape or barrel shape distortion is known. Conversely, image data (computer graphics) without distortion generated by the GPU may be displayed on a display with distortion, for example, a transparent head-up display. As described above, when an image with distortion remaining obtained from the optical lens is output to the output display 300 such as a flat display, or the image data without distortion obtained by the GPU is output to a display with distortion. When output to the output device 300, the image is projected and contracted according to the distortion. Therefore, the image processing apparatus 100 according to the present invention has a function of performing correction processing for restoring image data including such distortion to an original flat shape, or conversely, correcting processing for distorting flat image data.

  The image processing apparatus 100 receives the image data input from the input device 200 at the image data input unit 20. The image data input unit 20 temporarily stores the received image data in the line buffer 30 inside the apparatus. Further, the image data input unit 20 may develop the received image data in the external memory 30 outside the apparatus. The image data stored in the line buffer 30 (or the external memory 30) is appropriately read out by the image correction means 10 (particularly the coordinate conversion unit 13), and distortion correction is performed in units of pixels. Note that the image correction unit 10 may temporarily write the pixels for which the distortion correction processing has been completed into the line buffer 30. Thus, the line buffer 30 is used as a working memory for reading and writing data when the image correction unit 10 performs the distortion correction process. A specific configuration of the image correction unit 10 will be described later.

  The image data that has been subjected to the distortion correction processing by the image correction means 10 is sent to the filter unit 40. The filter unit 40 performs pixel interpolation processing in the same manner as the filter method performed in texture mapping while referring to the coordinate values of the corrected image data. For example, the filter unit 40 may perform color value interpolation, transparency interpolation, and blur value interpolation as pixel interpolation processing. Examples of filter methods used in pixel interpolation processing include bilinear filtering, trilinear filtering, anisotropic filtering, and bicubic filtering. Other known filter methods can also be employed.

  The image data that has undergone the interpolation processing in the filter unit 40 is transferred to an output device 300 such as an external memory or a display via the image output unit 50. The image data transferred to the output device 300 has been subjected to distortion correction processing in accordance with the shape of the display or the like. The image data after distortion correction can be directly displayed on the display as it is, or can be stored in an external memory and used for another processing thereafter.

  In the image processing apparatus 100 of the present invention, the filter unit 40 has an arbitrary configuration and can be omitted. That is, the image correction unit 10 may directly transfer the image data subjected to distortion correction to the output device 300 such as an external memory or a display.

  Next, the image correction unit 10 will be described in detail. As shown in FIG. 1, the image correction unit 10 includes a correction value storage unit 11, an LUT development unit 12, and a coordinate conversion unit 13. The correction value storage unit 11 stores an offset code table 11a and a compression LUT 11b. The LUT expansion unit 12 reads the compressed LUT 11b from the correction value storage unit 11 as necessary, and decompresses the compressed LUT 11b based on the offset code stored in the offset code table 11a to generate a correction LUT 11c. . The compression LUT 11b and the correction LUT 11c store substantially the same information and have a reversible relationship with each other. However, the information amount of the compression LUT 11b is smaller than the information amount of the correction LUT 11c. The coordinate conversion unit 13 reads image data stored in the line buffer 30 and the like, and performs coordinate conversion processing for distortion correction on a pixel basis based on the correction LUT 11c generated by the LUT development unit 12.

FIG. 2 shows an outline of a correction LUT used for coordinate conversion processing for distortion correction. The upper diagram in FIG. 2 shows the input image and the image after distortion, and the lower diagram in FIG. 2 shows an example of the correction LUT. As shown in FIG. 2, the coordinate value (x ₀ , y ₀ ) of the point P ₀ in the input image is changed to the coordinate value (x ₀ ′, y ₀ ′) of the point P ₀ ′ in the image after distortion. Need to convert. Therefore, by recording such pixel coordinate information before and after conversion in association with the correction LUT, the correction LUT can be used when creating a distorted image from the input image. The lower diagram of FIG. 2 shows information on the correction LUT (before compression). In the correction LUT, coordinate values before distortion correction and coordinate values after distortion correction are recorded in association with each other. By creating a coordinate list before and after conversion in advance as in this correction LUT, a post-distortion image can be created instantaneously using this information. However, such a correction LUT needs to record a coordinate value before conversion and a coordinate value after conversion for each pixel, and has a large amount of information (data amount). If such a correction LUT is stored in the memory as it is, it takes time to read out the correction LUT from the memory as necessary, and this compresses valuable memory space. Therefore, the present invention generates a compressed LUT by compressing the correction LUT, and stores the compressed LUT with a small amount of information in the memory. Then, if necessary, the compression LUT is decompressed and the correction LUT is expanded on the memory, and this correction LUT is used for image data distortion correction processing.

  FIG. 3 shows an example of an offset code table used for compression of the correction LUT. FIG. 4 shows a configuration example of a compressed LUT obtained by compressing the correction LUT and its compressed bits. 3 and 4 show an example in which the information amount per pixel of the correction LUT is compressed to an information amount of 16 bits / pixel in the compression LUT.

  As shown in FIG. 3, the offset code table records an offset value (difference value) between the coordinate value before distortion correction and the coordinate value after distortion correction in association with a predetermined offset code. For example, in this example, in the content part of the offset code table, the value of the integer part of the coordinate before and after conversion is compared for each of the X coordinate and the Y coordinate, and the offset value (difference value) of the value is recorded as information. Has been. In the offset code table, a predetermined offset code is assigned to each offset value. For example, the offset code is set as 4-bit information. However, the size of the offset code is not limited to 4 bits, and can be adjusted to an optimal size as appropriate. In addition, information for invalidating the X coordinate, the Y coordinate, or both can be recorded in the content portion of the offset code table in association with the offset code.

As shown in FIG. 4, the correction LUT is compressed using the values of the offset code table shown in FIG. 3, and converted into a compressed LUT. That is, the compression LUT has a configuration in which an offset code is recorded in association with a coordinate value compressed by referring to an offset value corresponding to the offset code. More specifically, in the present embodiment, in the offset code table shown in FIG. 3, a predetermined value is used for the offset value between the integer part of the coordinate value before distortion correction and the integer part of the coordinate value after distortion correction. An offset code is assigned. For this reason, in the compression LUT shown in FIG. 4, the offset code is associated with the fractional part (x ′ _frac , y ′ _frac ) of the coordinate value obtained by subtracting the offset value corresponding to the offset code. Note that “frac” shown in the figure means a fraction part.

  The conversion method from the correction LUT to the compression LUT will be described more specifically with reference to FIGS.

[1] Looking at the correction LUT before compression, in the part for converting the coordinate value (x, y) = (0, 0) to the coordinate value (x ′, y ′) = (0, 0), the integer part The offset value is (+0, +0). The offset value (+0, +0) is defined as offset code = 3 in the offset code table. In the conversion from (x, y) = (0, 0) to (x ′, y ′) = (0, 0), there is no difference in the values after the decimal point. Therefore, in the compressed LUT, (x ′ _frac , y ′ _frac , offset code) = (0, 0, 3) is recorded.

[2] Looking at the correction LUT before compression, in the part for converting the coordinate value (x, y) = (1, 0) to the coordinate value (x ′, y ′) = (1.2, 0), The offset value of the integer part is (+0, +0). The offset value (+0, +0) is defined as offset code = 3 in the offset code table. Further, in the conversion from (x, y) = (1, 0) to (x ′, y ′) = (1.2, 0), the value after the decimal point has a difference of 0.2 only in the x coordinate value. It does not occur in the y coordinate value. Therefore, in the compression LUT, it is recorded as (x ′ _frac , y ′ _frac , offset code) = (0.2, 0, 3).

[3] Looking at the correction LUT before compression, the portion for converting the coordinate value (x, y) = (2, 0) to the coordinate value (x ′, y ′) = (3.0, 1.1) , The offset value of the integer part is (+1, +1). The offset value (+1, +1) is defined as offset code = 11 in the offset code table. In the conversion from (x, y) = (2, 0) to (x ′, y ′) = (3.0, 1.1), the value after the decimal point is a difference of 0.1 only in the y coordinate value. Has occurred and not in the x-coordinate value. Therefore, in the compressed LUT, (x ′ _frac , y ′ _frac , offset code) = (0.0, 0.1, 11) is recorded.

  As described above, the correction LUT is compressed using the value of the offset code table and converted into a compressed LUT. In the correction LUT, the coordinate values of the x-coordinate and the y-coordinate are recorded including the integer part and the decimal part before and after the conversion. On the other hand, in the compression LUT, the difference between the integer parts of the coordinate values before and after conversion is represented by a 4-bit offset code value, and only the value of the decimal part of the converted x coordinate and y coordinate is the offset code value. Correspondingly recorded. Therefore, it can be said that the information amount of the compression LUT is significantly smaller than the information amount of the correction LUT.

  FIG. 4 shows a configuration example of compressed bits in the compressed LUT. The compression bit is, for example, 6 bits (lower bits) from 0 to 5 as data of the decimal part of the converted x coordinate, and 6 bits (lower bits) from 6 to 11 are converted to the decimal part of the y coordinate after conversion. 4 bits (upper bits) from 12 to 15 are used as offset code values. In this way, the data can be configured as a total of 16 bits. That is, in the compression LUT, the information amount per pixel used for distortion correction can be compressed to 16 bits. By compressing the amount of information in the LUT used for distortion correction in this way, information used in other processing (for example, mask information related to clipping, image ID, Information for brightness adjustment) can be stored.

  In the present embodiment, an example of information of 16 bits / pixel is shown as an example. However, regarding the size of each bit field (x coordinate, y coordinate, offset code) constituting the compressed bit and the size of the entire compressed bit, it is possible to select an optimum size for each application.

  As shown in FIG. 1, the compressed LUT 11b created as described above is stored in the correction value storage unit 11 together with the offset code table 11a. In FIG. 1, the correction value storage unit 11 is drawn as if constructed by one storage device (memory), but the correction value storage unit 11 is constructed by a plurality of storage devices (memory). Also good. For this reason, the offset code table 11a and the compression LUT 11b are not necessarily stored in the same storage device, and may be physically stored in separate storage devices.

  The LUT expansion unit 12 performs a correction LUT 11c based on the compression LUT 11b and the offset code table 11a stored in the correction value storage unit 11. The LUT development unit 12 may develop the correction LUT 11c into the correction value storage unit 11 (memory), may develop into other internal memory, or may develop into an external memory. In addition, the LUT expanding unit 12 can expand the correction LUT 11 c to the line buffer or the external memory 30. Specifically, as described above, the x-coordinate fractional part, the y-coordinate fractional part, and the offset code are recorded in association with each other in the compression LUT 11b. The LUT expansion unit 12 refers to the offset code table 11a based on the offset code in the compression LUT 11b, and reads the offset value (the integer part of the coordinate) of the coordinate associated with the offset code. Then, the LUT expansion unit 12 performs a calculation process of adding the coordinate offset value (integer part) read from the offset code table 11a to the decimal part of the x coordinate and the decimal part of the y coordinate. As a result, the compression LUT 11b can be decompressed based on the value of the offset code table 11a to generate the correction LUT 11c.

  When necessary image input is completed, the control unit 60 instructs the start of distortion correction processing, and the LUT development unit 12 reads the compressed LUT 11b from the correction value storage unit 11, decompresses it, and develops the correction LUT 11c. At the same time, the value of the correction LUT 11 c is input to the coordinate conversion unit 13. The coordinate conversion unit 13 refers to the correction LUT 11c generated by the LUT development unit 12, and performs distortion correction processing in units of one pixel on the image data stored in the line buffer 30 (or external memory). Run. That is, the coordinate conversion unit 13 refers to the correction LUT 11c and performs an operation of converting the coordinate value of each pixel of the input image before correction into the coordinate value after distortion correction. When the distortion correction processing is completed for each pixel, the coordinate conversion unit 13 sends the pixel information together with the corrected coordinate information to the filter unit 40.

In addition, the coordinate conversion unit 13 can also perform a distortion correction formula calculation in the distortion correction processing of image data. Specifically, the coordinate conversion unit 13 calculates the following expressions (1) and (2).
(1) x _d = x value obtained by the distortion correction formula + x value of the developed correction LUT (2) y _d = y value obtained by the distortion correction formula + y value of the developed correction LUT As the distortion correction formula here, a known correction formula can be used. For example, a known Brown-Conrady model may be used as a distortion correction formula. More specifically, the coordinate conversion unit 13 is implemented with an algorithm for calculating the following mathematical formula.

  When the calculation based on the above equation is completed, the coordinate conversion unit 13 inputs pixel information to the filter unit 40 together with the corrected coordinate information. As described above, the filter unit 40 performs pixel interpolation processing using a bilinear filter or the like in the same manner as the filter method performed in texture mapping while referring to the corrected coordinate values. Thereafter, the result calculated by the filter unit 40 is stored in a display or a memory for storing output image data through the image output unit 50. The same calculation is performed for the image output resolution, and a distorted image is generated.

In the example shown in FIG. 2 and FIG. 4 of the present application, the distortion correction formula is not applied to the coordinate value before conversion in the correction LUT, but the distortion correction formula is combined with the coordinate value before conversion. It is also possible. In that case, the value of (x _d , y _d ) to which the above distortion correction formula is applied is stored as the coordinate value before conversion in the correction LUT, and (x _d ′, y _d ′) is stored as the coordinate value after conversion. Will be stored. Even in such a case, the subsequent generation of the compression LUT and the decompression process thereof can be performed in the same manner as described above.

  Next, a processing flow of the control unit 60 in the image processing apparatus 100 will be described with reference to FIG. The control unit 60 first performs preprocessing for creating a compressed LUT before starting input of image data (step S1). When an input device 200 such as a camera input device is connected to the image processing device 100, the control unit 60 detects distortion of the optical lens included in the input device 200 by image processing or 3D measurement using a checker board, and a distortion map is obtained. create. The distortion map of the optical lens is a map showing how much distortion is generated in which region of the image input from the optical lens with respect to a flat image without distortion. On the other hand, when an output device 300 such as a transmissive head-up display is connected to the image processing apparatus 100, the control unit 60 may detect distortion on the head-up display and create a distortion map of the display. The display distortion map is a map that shows how much distortion the display screen of the head-up display produces in a flat image without distortion. In this sense, it can be said that the control unit 60 of the image processing apparatus 100 includes a distortion detection unit of the input device 200 or the output device 300. Further, the control unit 60 creates a correction LUT before compression based on the distortion map. Further, the offset value table described above has already been created and stored in the correction value storage unit 11. Therefore, the control unit 60 creates a compressed LUT by compressing the correction LUT based on the value of the existing offset code table, and stores it in the correction value storage unit 11 together with the offset code table. The method for creating the compressed LUT from the correction LUT is as described above.

  Thereafter, input of image data is started to the image processing apparatus 100 (step S2). When image data is input, a pixel image necessary for processing is prepared, and then a loop for output pixels is entered. In the loop, first, the corresponding coordinate-converted value is expanded from the compression LUT to the correction LUT according to the coordinates of the input pixel, and coordinate conversion is performed based on the correction LUT (step 3). . Thereafter, necessary image data is read from the line buffer or the external memory 30 to prepare output pixels and input to the filter unit 40 (step S4). Thereafter, after filter processing such as bilinear filtering is performed in the filter unit 40, the image after the filter processing is sent to the image output unit 50, and the image data after distortion correction is stored in the external memory through the image output unit 50. Or display on the display (step S5). Similar processing is performed by looping the image output resolution, and a distorted image is generated.

  Next, an example of the expanded contents of the compression table will be described with reference to FIG. As described with reference to FIG. 4, the compression bits per pixel constituting the compression table allocate the decimal part of the x coordinate after conversion to the lower 6 bits from 0 to 5, and to the lower 6 bits from 6 to 11 The decimal part of the converted y-coordinate is assigned, and the offset code value is assigned to the upper 4 bits from 12 to 15, and the total is 16 bits. As described above, since the compressed bit is data with a small amount of information, a surplus space is generated in the memory for storing the compressed bit. Therefore, the compression bit can be expanded to store other useful information.

  For example, as shown in the upper diagram of FIG. 6, the compression bit is expanded to 32-bit data of 0 to 31, and an expansion code is assigned to 16 bits of 16 to 31. The extension code indicates processing information used for image processing for each pixel, and an example of definition of the extension code is shown in the lower diagram of FIG. For example, processing information instructed by the extension code includes mask information related to clipping (the presence or absence of output can be controlled for each pixel) and image ID (for stitching. When displaying multiple input images in combination, Each input image can be selected), or brightness adjustment information (in-vehicle use, etc., depending on the time zone and light ON / OFF status, visibility can be increased by changing the color and brightness Possible). For example, as shown in the lower diagram of FIG. 6, processing information of a clipping mask (transparency setting) can be assigned to 16 to 19 bits, an image ID can be assigned to 20 to 23 bits, and a luminance offset can be assigned to 24 to 31. As described above, by assigning processing information useful for processing in units of pixels to the extended portion of the compressed bit, it is possible to realize the efficiency and speed of the image drawing processing.

  Next, another compression method of the correction LUT will be described with reference to FIG. FIG. 7 shows an example in which the information amount of the correction LUT can be halved when the shape of the image after distortion correction is symmetrical. As shown in FIG. 7, when the image shape after correction is symmetrical, for example, the correction LUT is created only from the left half of the converted image (from x = 0 to the image center (width / 2)). If so, the right half of the converted image can be omitted. That is, for the right half after conversion, a simple calculation of (width-x) can be performed, and the calculated value can be used as the x value of the right half. Although FIG. 7 illustrates the case where the converted image is bilaterally symmetric, the amount of information in the correction LUT can be halved in the same manner when the converted image is vertically symmetric.

  As mentioned above, in this specification, in order to express the content of this invention, embodiment of this invention was described, referring drawings. However, the present invention is not limited to the above-described embodiments, but includes modifications and improvements obvious to those skilled in the art based on the matters described in the present specification.

  The present invention relates to an image processing apparatus and a program having a function for correcting distortion of image data. The apparatus and program according to the present invention can be applied to a computer for image processing, a digital camera, a head-up display, or the like. Therefore, the present invention can be suitably used in computer related industries, game related industries, medical related industries, and in-vehicle related industries.

DESCRIPTION OF SYMBOLS 10 ... Image correction means 11 ... Correction value memory | storage part 11a ... Offset code table 11b ... Compression LUT
11c ... LUT for correction 12 ... LUT development unit 13 ... Coordinate conversion unit 20 ... Image data input unit 30 ... Line buffer or external memory 40 ... Filter unit 50 ... Image output unit 60 ... Control unit 100 ... Image processing device 200 ... Input device 300 ... Output device

Claims (6)

An image processing apparatus comprising image correction means (10) for correcting distortion of image data,
The image correction means (10)
A correction value storage unit (11), a lookup table (LUT) development unit (12), and a coordinate conversion unit (13);
The correction value storage unit (11)
An offset code table (11a) in which an offset value between a coordinate value before distortion correction and a coordinate value after distortion correction is associated with a predetermined offset code;
A compressed LUT (11b) in which the offset code is associated with a coordinate value compressed with reference to an offset value corresponding to the offset code;
The LUT expansion unit (12)
Based on the offset code table (11a) and the compression LUT (11b), a correction LUT (11c) in which the coordinate value before distortion correction and the coordinate value after distortion correction are associated is generated,
The coordinate conversion unit (13)
An image processing apparatus for correcting a coordinate value of the image data based on the correction LUT (11c). The offset code table (11a) associates an offset value between the integer part of the coordinate value before distortion correction and the integer part of the coordinate value after distortion correction with a predetermined offset code,
The image processing apparatus according to claim 1, wherein the compression LUT (11b) associates the offset code with a fractional part of coordinate values obtained by subtracting an offset value corresponding to the offset code. The compressed LUT (11b) is associated with an extended code indicating processing information used for image processing at the coordinate value in addition to the compressed coordinate value and the offset code. The image processing apparatus according to claim 2. When the image data after distortion correction is vertically symmetric or horizontally symmetric, the correction LUT (11c) holds coordinate values of only one of the upper and lower sides or only one of the left and right sides with the symmetry axis as a boundary. 4. The image processing device according to any one of 3. The coordinate conversion unit (13) is obtained by applying a predetermined distortion correction formula to the coordinate values after distortion correction held in the correction LUT (11c) and the coordinate values of the image data. The image processing device according to claim 1, wherein the coordinate value of the image data is corrected based on the coordinate value. A program for causing a computer to function as an image processing apparatus including image correction means (10) for correcting distortion of image data,
The image correction means (10)
A correction value storage unit (11), an LUT expansion unit (12), and a coordinate conversion unit (13);
The correction value storage unit (11)
An offset code table (11a) in which an offset value between a coordinate value before distortion correction and a coordinate value after distortion correction is associated with a predetermined offset code;
A compressed LUT (11b) in which the offset code is associated with a coordinate value compressed with reference to an offset value corresponding to the offset code;
The LUT expansion unit (12)
Based on the offset code table (11a) and the compression LUT (11b), a correction LUT (11c) in which the coordinate value before distortion correction and the coordinate value after distortion correction are associated is generated,
The coordinate conversion unit (13)
A program for correcting the coordinate value of the image data based on the correction LUT (11c).

Images