JP6040217B2 - Projection apparatus and projection method - Google Patents

Description

The present invention relates to an apparatus for projecting an image by applying trapezoidal correction.

  Projection devices such as liquid crystal projectors can be used to move upward or downward from a position that does not face the projection surface, such as a screen, when the installation location is limited or when the projector is installed so as not to obstruct the viewer's field of view. May have to be projected. In such a case, the angle formed between the optical axis and the projection plane is different from a preset angle that can be appropriately projected, and thus the projected image projected onto the projection plane is distorted. For example, when a rectangular image is displayed, a trapezoidal distortion (keystone distortion) occurs.

  A technique for correcting a projected image into a rectangular shape by performing a trapezoidal correction (trapezoidal distortion correction, keystone correction) on the projected image with the distortion of the projected image caused by the projection method on the projection surface in this way is known. It has been. Some projection apparatuses have a function of detecting distortion of an image on the projection surface from the tilt angle of the projection apparatus and the shape of the projected image, and automatically correcting the trapezoid.

  In recent years, with the start of digital television broadcasting and the like, the resolution (number of pixels) of video content has increased, and there is an increasing demand for viewing high-resolution video content using a projection device. In order to project high-resolution video content in the projection device, the time taken for one frame of the video content is constant. Therefore, as the number of pixels increases, the clock for processing one pixel is increased in speed. Reproduction of pixel video content was realized.

  However, when keystone correction is performed while projecting high-resolution video content with a projection apparatus, multiplication processing and memory access related to keystone correction are required, and thus there are the following problems. That is, when a high-speed clock is required to reproduce multi-pixel video content, multiplication processing and memory access related to trapezoid correction cannot be completed within one frame, and the video frame rate may fail.

  Patent Document 1 discloses a technique for preventing the frame rate from failing by reducing the number of bits of color information of video for multi-pixel video content as described above.

JP 2007-251723 A

  However, in Patent Document 1, since the number of bits of color information is reduced, image quality deterioration occurs in a projected image. Further, in the trapezoidal correction process accompanied with the image enlargement process, image quality deterioration due to the reduction in the number of bits of the color information may appear remarkably.

  In addition, it is possible to reduce the time required for processing video content by dividing multi-pixel video content and performing parallel keystone correction processing using multiple image processing circuits. It is done.

  For example, when the projection device projects a rectangular image as shown in FIG. 3A on the screen as the projection surface from the lower side of the screen, the projected image projected on the screen is shown in FIG. ) To form a trapezoidal shape. When the rectangular image is not divided, the rectangular image obtained by correcting the distortion on the screen as shown in FIG. 3D by deforming and projecting the rectangular image into a trapezoidal shape as shown in FIG. A similar image can be projected. At this time, if the projected images are classified into the areas D1 to D4 divided into four equal parts in the vertical direction as shown in FIG. 3A, the respective areas are different in size as shown in FIG. It is transformed into a trapezoidal shape. Specifically, since the reduction ratios in the horizontal and vertical directions change depending on the optical distance to the projection surface in each area, the areas become smaller in order from D4 to D1 after the keystone correction.

  When the keystone correction is performed by dividing the projection image into a plurality of divided images, the correction image obtained by performing the keystone correction from each of the divided images is reduced and deformed in the horizontal and vertical directions. Will be included. In other words, the circuit for applying the keystone correction includes a memory having a capacity capable of storing the divided images. However, when the correction image newly obtained by applying the keystone correction is stored in the memory, the circuit is reduced in the horizontal and vertical directions. Therefore, the memory includes the address of the line where the pixel value does not exist.

  Therefore, in order to project a plurality of corrected images obtained by trapezoidally correcting each of the plurality of divided images as one connected trapezoidal image as shown in FIG. 3C, the following processing is performed. There is a need. That is, after reading out a plurality of corrected images from the memory, a process of removing a line having no pixel value and synthesizing it into one connected trapezoidal image is necessary as compared with a case where a rectangular image to be projected is not divided. It can be a problem.

The present invention, by performing the keystone correction process in parallel by a plurality of correction means, an object and Turkey to reduce the time required for processing.

The projection apparatus according to the present invention includes a first image including a first region from a predetermined image based on a correction amount of a trapezoid correction process performed on an image projected on a projection plane, and the first image Generating means for generating a second image including a second area having a size different from the size of the area; first correcting means for performing a first trapezoidal correction process on the first image; and A second correction unit that performs a second trapezoidal correction process on the second image; an image that has undergone the first trapezoidal correction process; and an image that has undergone the second trapezoidal correction process. A projection unit that projects an image generated based on the projection plane; a clock that is provided in common to the first correction unit and the second correction unit; and the first trapezoid correction process. And providing a clock for controlling execution of the second trapezoidal correction process Possess a control means, wherein the control means is fast enough target keystone correction process is large, and the clock based on either the larger image of the first image and the second image, the Control is provided to provide the first correction means and the second correction means .

According to the present invention, by performing the keystone correction process in parallel by a plurality of correction means allows the Turkey to reduce the time required for processing.

FIG. 2 is a block diagram showing a functional configuration of the liquid crystal projector according to the embodiment. The block diagram which shows the function structure of the image division part which concerns on embodiment. The figure for demonstrating the trapezoid correction process of a perpendicular direction. FIG. 5 is a diagram for explaining writing of a projection image in an image dividing unit according to the first embodiment. FIG. 3 is a block diagram showing a functional configuration of an image processing unit according to the embodiment. FIG. 4 is a diagram for explaining a projection image dividing method according to the first embodiment. FIG. 4 is a diagram illustrating an example of a corrected image output from each image processing unit according to the first embodiment. 6 is a flowchart of trapezoid correction processing according to the first embodiment. FIG. 10 is a diagram for explaining writing of a projection image in an image dividing unit according to the second embodiment. FIG. 6 is a diagram for explaining a projection image dividing method according to the second and third embodiments. FIG. 6 is a diagram for explaining the correspondence between divided images and corrected images according to the second and third embodiments. 6 is a timing chart when dividing the projection images according to the second and third embodiments. 10 is a flowchart of trapezoid correction processing according to the second embodiment. 10 is a flowchart of trapezoid correction processing according to the third embodiment.

(Embodiment 1)
Hereinafter, a preferred embodiment of the present invention will be described in detail with reference to the drawings. In the embodiment described below, an example in which the present invention is applied to a liquid crystal projector that can divide a projection image and apply a trapezoid correction process to each of the divided images as an example of a projection device. Will be explained.

  FIG. 1 is a block diagram showing a functional configuration of a liquid crystal projector 100 according to an embodiment of the present invention.

  The control unit 101 is, for example, a CPU, and develops a control program for each block of the liquid crystal projector 100 stored in a non-illustrated non-volatile memory in a RAM (not shown) to execute the control program. Control the operation of each block. The operation unit 102 is an input interface that receives an input from the user such as a power button or a cursor key, and transmits an operation performed by the user to the control unit 101. The first infrared receiver 121 and the second infrared receiver 122 are blocks that receive infrared signals provided on the front and rear surfaces of the liquid crystal projector 100. Similar to the operation unit 102, the first infrared reception unit 121 and the second infrared reception unit 122 analyze the infrared signal transmitted when the user operates the remote controller, and perform an operation performed by the user on the control unit 101. To transmit. The display unit 127 is a display device such as a small LCD, for example, and is controlled by the display control unit 128 to display a notification of setting of the liquid crystal projector 100, GUI data, and the like to the user.

  The analog input unit 110, the digital input unit 112, the USB I / F 113, the card I / F 114, the communication unit 115, and the built-in memory 116 are all interfaces that the liquid crystal projector 100 includes and receives video signal inputs. When an analog video signal is input to the analog input unit 110, the input analog video signal is changed to a digital video signal by the A / D conversion unit 111 and then input to the image dividing unit 133. Further, not only video signals but also image files, moving image files, and the like are input from each input interface, converted into video signals that can be played back by the liquid crystal projector 100 by the file playback unit 132, and then input to the image division unit 133. Is transmitted.

  The image dividing unit 133 divides one image (projected image) related to one frame of the input video signal into a plurality of images (divided images) and outputs the images. In the present invention, when various image processes are applied with only one image processing unit, the process takes time. Therefore, a high-resolution image (the number of pixels is not sufficient to achieve a specified number of frames / second). (Assuming many images). Therefore, the liquid crystal projector 100 according to the present embodiment includes four image processing units, a first image processing unit 117, a second image processing unit 118, a third image processing unit 119, and a fourth image processing unit 120, and an image dividing unit. 133 divides the projected image into four regions and outputs them to each image processing unit. The timing for presenting the image is based on the vertical synchronization signal (VSYNC) and horizontal synchronization signal (HSYNC) received together with the input video signal.

  The first image processing unit 117, the second image processing unit 118, the third image processing unit 119, and the fourth image processing unit 120 each presents an input image on the liquid crystal unit 104 to be described later. Various image processes are applied and output to the liquid crystal drive unit 105. For example, resolution conversion processing for converting the input image into the display resolution of the liquid crystal unit 104, tone correction, gamma correction processing, and the like are performed. In addition, when the liquid crystal unit 104 performs AC driving, an image can be displayed in the liquid crystal unit 104 regardless of the voltage direction by outputting the input video signal by doubling the number of frames. In the present invention, when an image is projected from an oblique direction on the screen in each image processing unit, when the projected image is distorted, for example, in a trapezoidal shape, the shape of the image is deformed so as to cancel the distortion. Performs keystone correction (keystone correction). The trapezoidal correction process is performed by, for example, grasping the tilt angle of the liquid crystal projector 100 detected by the tilt sensor 134 or the distortion of the shape of the predetermined pattern projected on the projection plane when the imaging unit 124 is provided. This is done by determining the amount of trapezoidal correction by detecting the shape of the screen, such as the edge of the screen. In the trapezoidal correction process, it is possible to perform a highly accurate trapezoidal correction process by using not only the information on the tilt angle but also the zoom state and optical characteristic information of the projection optical system 107, and the control unit 101 can store these pieces of information. It is provided to each image processing unit together with the coordinate information of the image end point after the keystone correction. For example, when the user operates the operation unit 102 to set a correction amount for arbitrary trapezoidal correction, each image processing unit performs a keystone correction process based on the set value. In the trapezoidal correction process, the trapezoidal distortion on the projection surface is canceled by changing the reduction ratio with respect to at least one of the horizontal and vertical directions of the image formed on the liquid crystal unit 104, which will be described later. The image can be brought close to a rectangular image having a normal aspect ratio.

  The liquid crystal driving unit 105 synthesizes an image to which various image processing is applied in each image processing unit into one image form and outputs it to the liquid crystal unit 104. Specifically, in the liquid crystal drive unit 105, for example, the pixel signals read out and input from the internal memory of each image processing unit in the order of the addresses of the internal memory are input into the liquid crystal so that the composite image to be projected is in the raster scan order. The data is output to the unit 104 and displayed. The liquid crystal unit 104 is, for example, one full-color liquid crystal element or RGB liquid crystal element, and the liquid crystal driving unit 105 changes the voltage output to each pixel based on the input pixel signal, so that the liquid crystal A projected image (corrected image) is formed on the element. The image formed on the liquid crystal unit 104 is imaged (projected) on an external screen (projection surface) via the projection optical system 107 by the light source 106 driven by the light source control unit 108. The light amount of the light source 106 is controlled by the light source control unit 108 according to the light amount value calculated so that the luminance level on the projection surface measured by the screen photometry unit 125 is appropriate, for example. Information on the amount of light from the light source 106 is detected by a sensor of the light source photometry unit 126 provided on the optical path and fed back to the light source control unit 108. The projection optical system 107 is a lens group including a zoom lens, a focus lens, a shift lens, and the like. The projection optical system 107 is driven by the optical system control unit 109 to change the zoom magnification of the projected image, adjust the focus, control the projection position, and the like. Is possible. For example, the focal length is calculated from the distance between the screen and the liquid crystal projector 100 detected by the focus detection unit 123, and the control unit 101 sends the projection optical system 107 to the optical system control unit 109 based on the calculated focal length. Drive the focus lens.

  The power input unit 130 is an interface that receives an input of AC power from the outside, rectifies the input AC power into a predetermined voltage, and supplies the rectified voltage to the power source unit 103. The power supply unit 103 is a block that supplies power to each block of the liquid crystal projector 100 that performs power supply drive via the control unit 101. The power supply unit 103 stores power in the battery 129 by supplying power to the battery 129, and can use the battery 129 as backup power even when AC power is not input to the power input unit 130. . The timer 131 is a timer that detects an operation time or the like related to processing of each block of the liquid crystal projector 100.

  In the liquid crystal projector 100 of the present embodiment having such a configuration, one image to be projected is divided into a plurality of images, and after performing keystone correction processing on each of the images, the image is projected as one image to which the keystone correction is applied. The overall process when doing this will be described in detail.

  In the present embodiment, keystone correction having a size obtained by dividing the projection image into four equal parts from each of the first image processing unit 117, the second image processing unit 118, the third image processing unit 119, and the fourth image processing unit 120 is performed. The applied corrected image is output to the liquid crystal drive unit 105. The corrected image output from each image processing unit to the liquid crystal driving unit 105 is an image that does not require processing for combining a plurality of corrected images into one connected image in consideration of image discontinuity due to a line having no pixel value. It has become. Specifically, it is an image obtained by dividing the rectangular image including the line having no pixel value, which is formed in the liquid crystal unit 104 by performing the keystone correction into four equal parts.

  Specifically, the corrected image formed in the liquid crystal unit 104 by performing trapezoidal correction on the projection image as shown in FIG. 3A is as shown in FIG. 3C, and is output from each image processing unit. The corrected image is as shown in FIG. FIG. 7 shows a corrected image output from each image processing unit, which is an image obtained by dividing a rectangular image finally formed in the liquid crystal unit 104 into four equal parts in the vertical direction. That is, when a corrected image as shown in FIG. 7 is generated in each image processing unit, the liquid crystal driving unit does not perform a process of combining a plurality of corrected images into one connected image in which lines having no pixel values are sorted. It is possible to output a connected image to 105. That is, sequentially from the first image processing unit 117, the corrected images obtained by the respective image processing units are read in the order of addresses and output to the liquid crystal driving unit 105, whereby the connected images are input to the liquid crystal driving unit 105 in the raster scan order. The same effect as that. That is, in order from the first image processing unit 117, the corrected images obtained by the respective image processing units are read out in the same address order as that in the case of projecting without performing the trapezoidal correction process, thereby performing a raster scan of the connected images. Data can be read out in the same order of addresses as before.

  Hereinafter, specific processing of the image dividing unit 133 will be described with reference to the block diagram of FIG. 2 showing the internal configuration of the image dividing unit 133.

  An image related to one frame of the video signal input from an input interface such as the digital input unit 112 is input to the image dividing unit 133 by the control unit 101. In the present embodiment, the image dividing unit 133 first divides an input image into four equally sized areas in the vertical direction as shown in FIG. Therefore, the image dividing unit 133 includes a first divided memory 201, a second divided memory 202, a third divided memory 203, and a fourth divided memory 204 that store the divided images as shown in FIG. The image dividing unit 133 also manages a storage address of the divided image and controls input / output timing to each image processing unit, a first divided memory control unit 205, a second divided memory control unit 206, A third divided memory control unit 207 and a fourth divided memory control unit 208 are provided.

  One image input to the image dividing unit 133 is divided by writing an image of an area to be divided into each divided memory according to an instruction from the control unit 101. The image input to the image dividing unit 133 is read in the scanning order according to the scanning direction of the raster scan, for example. Therefore, when the image is divided into four in the vertical direction as in the present embodiment, VSYNC input together with the video signal is used. . In the present embodiment, the scanning direction of the raster scan is assumed to be constant in the horizontal direction (from left to right). After reading (scanning) the pixels in the horizontal line from the upper left pixel in the image, the lower line is scanned. It is assumed that scanning in the horizontal direction is repeated from the leftmost pixel. The first divided memory control unit 205, the second divided memory control unit 206, the third divided memory control unit 207, and the fourth divided memory control unit 208 generate a pulse in a period of one quarter of the input VSYNC. Generate QVSYNC. For example, in the case of an image of horizontal Ht pixels and vertical Vt pixels as shown in FIG. 4A, VSYNC generates a pulse from the upper left pixel of the image every time a Vt line is read in the vertical direction, so QVSYNC is Occurs every time a Vt / 4 line is read. Then, as shown in the time chart of FIG. 4B, each divided memory control unit transmits a write permission signal, a chip select (CS) signal, and write address information to the divided memory using different QVSYNCs. As a result, the image signal input in the raster scan order is written into the divided memory in which writing is permitted for each line in the vertical direction, and thus is divided into four divided memories. That is, D1 in FIG. 4A is written in the first divided memory 201, D2 is written in the second divided memory 202, D3 is written in the third divided memory 203, and D4 is written in the fourth divided memory 204.

  The image signals thus read into the respective divided memories and divided are further controlled by the first selector 209, the second selector 210, the third selector 211, and the fourth selector 212, so that each image processing unit trapezoidal. A divided image for applying the correction process is output. In the present embodiment, each image processing unit 133 to each image processing unit so that the corrected image output from each image processing unit becomes an image (equal divided image) obtained by equally dividing the rectangular image formed on the liquid crystal unit 104. Controls the reading of each divided memory and the operation of each selector at the time of output. Specifically, for example, as shown in FIG. 6A, the operations of the divided memory and the selector are controlled so as to have a partially overlapping region (overlap region) between adjacent divided images.

  That is, the divided image output to each image processing unit includes an image having a predetermined number of lines from one of the image areas (D1 to D4) obtained by dividing the projection image into four equal parts and the area adjacent to the area. An area is included. In the example of FIG. 6A, the Vs line of the adjacent region is included as the overlap region in the region of Vt / 4 line obtained by dividing the projection image into four in the vertical direction. In this way, by projecting the projected image to each image processing unit in an area wider than the equally divided image, if the trapezoidal correction is performed, the adjacent equally divided address becomes a line where no pixel value exists. It is possible to store pixel value information obtained by performing keystone correction on an image.

  The information on the predetermined number of lines for determining the overlap region may be stored in, for example, a non-volatile memory (not shown), and the projection image is input to the image dividing unit 133 and the control unit 101 What is necessary is just to read and to transmit to the image division part 133. In the present embodiment, the information on the number of lines that define the overlap region is described below as being predetermined. However, the user may be able to set the information using the operation unit 102, for example. Further, in the present embodiment, an overlap area is included in the vertical direction of each divided image area. This is not only the case where the liquid crystal projector 100 performs projection from below the screen, but also from above the screen. This is because it is possible to cope with projection.

  The first selector 209, the second selector 210, the third selector 211, and the fourth selector 212 respectively convert the image read from the divided memory in accordance with the VSYNC signal as shown in FIG. Output to the processing unit. At this time, the control unit 101 controls the respective divided memory control units and selectors so as not to simultaneously read from the respective divided memories.

  Note that the read clock from the divided memory may be slower than the write clock. When the read clock speed is set to a quarter of the write clock speed, all data is read from four memories in one frame period. In the case of the present embodiment, in addition to all the data in the first divided memory 201 to the fourth divided memory 204, a maximum of 2 Vs lines for the overlap region is sent to the first selector 209 to the fourth selector 212. For this reason, when reading is performed at a speed equal to or lower than a quarter of the writing clock, data reading is not completed in one frame period, which causes processing delay and image update failure. That is, if the memory read clock speed is set to be higher than one quarter of the write clock speed, data read can be completed in one frame.

  The four image signals divided by the image dividing unit 133 are subjected to image processing in parallel by the four image processing units. Since the first image processing unit 117, the second image processing unit 118, the third image processing unit 119, and the fourth image processing unit 120 all have the same configuration, hereinafter, the first image processing unit 117 is taken as an example. Processing in the image processing unit will be described.

  FIG. 5A is a block diagram illustrating a functional configuration of the first image processing unit 117.

  The image signal input to the first image processing unit 117 is stored in the designated address of the first image processing memory 506 by the first image processing memory control unit 505, and then the resolution conversion unit 501 is instructed by the control unit 101. And resolution conversion processing is applied. The resolution conversion unit 501 converts the input image into a predetermined resolution in accordance with the setting of the liquid crystal unit 104, and is stored again at the designated address of the first image processing memory 506 by the first image processing memory control unit 505. The In the present embodiment, it is assumed that the first image processing memory 506 has a minimum capacity necessary for performing various image processes in the first image processing unit 117. That is, the image output from the resolution conversion unit 501 is stored from the top address of the first image processing memory 506, and is overwritten and stored on the address where the image input to the first image processing unit 117 is stored. . However, if the first image processing memory 506 has a sufficient area, the image to which the resolution conversion process is applied and the image input to the first image processing unit 117 are stored separately at addresses and can coexist. It may be.

  A trapezoid correction process is applied to the image whose resolution has been converted by the resolution conversion unit 501 by the trapezoid correction unit 502. Here, the functional configuration of the trapezoidal correction unit 502 will be described in more detail with reference to the block diagram of FIG.

  The trapezoidal correction memory control unit 511 is a block that controls processing of the trapezoidal correction unit 502. The trapezoid correction memory control unit 511 receives an instruction from the control unit 101, acquires the image to which the resolution conversion is applied by the resolution conversion unit 501 and applies the keystone correction processing from the first image processing memory 506, and stores it in the block memory 512. Remember. At this time, the pixel information read by the trapezoidal correction memory control unit 511 and the address information of the block memory 512 in which the read pixels are stored are determined by a coordinate calculation unit 514 described later.

  The coordinate calculation unit 514 receives from the control unit 101 the tilt angle, the zoom state of the projection optical system 107, the optical characteristic information, and the coordinate values of the end points of the four corners of the trapezoidally corrected image from the control unit 101. Then, the coordinate calculation unit 514 calculates coordinate correspondence information indicating the correspondence between the coordinates of the image after the keystone correction and the image after the keystone correction based on the received information. The coordinate correspondence information is information on the pixel stored at which address, among the divided images stored in the first image processing memory 506, the pixel information at an arbitrary coordinate after the keystone correction. Is information. In the present embodiment, a method for performing trapezoid correction by extracting and arranging information of pixels after trapezoid correction from pixels of an image before trapezoid correction will be described. The color information of the corrected pixel may be calculated from the color information of the plurality of pixels before correction. The coordinate correspondence information calculated in this way is transmitted to the trapezoid correction memory control unit 511, and the keystone correction memory control unit 511 reads image information from the first image processing memory 506 based on the coordinate correspondence information, and the block memory 512 is stored. At this time, it is assumed that the image information after the keystone correction is stored in the block memory 512 in the order of raster scanning.

  The filter processing unit 513 is a two-dimensional filter that performs, for example, enlargement / reduction processing and removal of aliasing distortion associated with the trapezoid correction processing on an image to which the trapezoid correction processing stored in the block memory 512 is applied. After applying the interpolation process to the image stored in the block memory 512, the filter processing unit 513 outputs the obtained image to the first image processing memory 506 again so as to be arranged at addresses in the order of raster scanning. Note that the address at which the image to which the keystone correction processing is applied to the first image processing memory 506 is stored is controlled by the first image processing memory control unit 505. In addition, when the image reduction process is performed by the filter processing unit 513, a non-signal portion that is a pixel having no pixel value is generated. Therefore, a mask signal such as blanking is applied to the non-signal portion using a masking circuit (not shown). May be added.

  In the present embodiment, when the trapezoid correction process in the trapezoid correction unit 502 is applied and a corrected image is written in the first image processing memory 506, an image (extracted image) that is finally output from the first image processing unit 117 is used. ) Only information is extracted and written. That is, since each image processing unit is input with a divided image including information on the overlap region, the corrected image obtained by performing the keystone correction process in each image processing unit includes an overlapping image. become. For this reason, the trapezoidal correction memory control unit 511 extracts an image finally output from each image processing unit with reference to the coordinate correspondence information when the interpolation processing in the filter processing unit 513 is completed, and performs image processing. Write to memory. For example, in the trapezoidal correction unit 502 of the second image processing unit 118, information on the image included in the region from the Vt / 4 + 1 line to the Vt / 2 line of the image output to the liquid crystal driving unit 105 and formed on the liquid crystal unit 104. Are extracted from the corrected image. In other words, the trapezoidal correction memory control unit 511 calculates the region of the Vt / 4 + 1 to Vt / 2 line from the coordinate correspondence information in the corrected image to which the interpolation processing in the filter processing unit 513 is applied, and only the region concerned is calculated. One image processing memory 506 is controlled to be stored.

  The size of the image output from the trapezoid correction unit 502 is most preferably a shape that is equally divided into four with respect to the panel. This is because the clocks for processing such as gamma correction and unevenness correction after the trapezoidal correction in the first image processing unit 117 can be lowered most by outputting in four equal parts. However, since the first image processing unit 117 to the fourth image processing unit 120 can output up to the input resolution including the overlap region, for example, the first image processing unit 117 outputs the Vt / 4-Vu line. The second image processing unit 118 may output an image for Vt / 4 + Vu lines. That is, in the present embodiment, only the case of equal division will be described, but it is only necessary to output the image corresponding to the resolution of the liquid crystal unit 104 from the four image processing units to the liquid crystal driving unit 105.

(Keystone correction processing)
A trapezoid correction process executed by the trapezoid correction unit 502 of this embodiment having such a configuration will be described with reference to the flowchart of FIG. In the trapezoidal correction process, the tilt angle of the liquid crystal projector 100 is detected by the tilt sensor 134, and the projected image is divided into divided images including the overlap region by the image dividing unit 133 and input to each image processing unit. In this case, it is assumed that the control unit 101 starts. However, the keystone correction process may be started by the control unit 101 before applying the keystone correction process in at least the image processing unit.

  In step S <b> 801, the control unit 101 acquires information on the current tilt angle of the liquid crystal projector 100 from the tilt sensor 134 and grasps the positional relationship between the liquid crystal projector 100 and the screen (angle detection). The control unit 101 calculates a necessary correction amount from the obtained tilt angle information so that the image projected on the screen is displayed in a similar shape to the projected image (S802). Specifically, for example, when the projected image is an image of Ht × Vt pixels, the control unit 101 displays the projected image similar to the projected image based on the obtained tilt angle information. The information of the four end points of the trapezoid as shown in (c) is calculated.

  In step S803, the control unit 101 calculates coordinate correspondence information representing coordinates after trapezoid correction processing for all pixels stored in the image processing memory from the correction amount information calculated in step S802. Then, the control unit 101 uses the coordinate correspondence information to output an image signal that is output from each image processing unit and included in an area obtained by equally dividing the image formed on the liquid crystal unit 104 into an image stored in each image processing memory. It is determined whether the signal is included after trapezoidal correction processing (S804). Specifically, for example, in the case of the second image processing unit 118, the control unit 101 first converts the coordinate information after the keystone correction processing of the pixels at the four end points of the image stored in the second image processing memory 507 into the coordinates. Calculate using correspondence information. Then, the control unit 101 outputs Vt / 2 from the Vt / 4 + 1 line, in which the vertical coordinates after the trapezoid correction processing of the four end points are regions of an image output from the second image processing unit 118 and formed on the liquid crystal unit 104. Determine whether the vertical coordinates of the line are included.

  When it is determined that the image area after the trapezoid correction processing includes an image to be output to the liquid crystal driving unit 105 in all the image processing units, the control unit 101 moves the process to S805. Then, the control unit 101 causes the trapezoid correction unit 502 of all the image processing units to perform the keystone correction process according to the coordinate correspondence information, extracts the image signal of the area output from each image processing unit, and stores it in the image processing memory. To complete the keystone correction process.

  Note that the corrected image stored in the image processing memory after the keystone correction process is applied is read out in the order of addresses of the image processing units so that the control unit 101 forms, for example, the raster scan order of the image formed on the liquid crystal unit 104. And input to the liquid crystal driving unit 105. Specifically, in the present embodiment, an image obtained by dividing an image including a line having no pixel value formed in the liquid crystal unit 104 into four in the vertical direction is stored in the image processing memory of each image processing unit. Therefore, the control unit 101 reads images from the first image processing memory 506 in the order of addresses and causes the liquid crystal driving unit 105 to output them. Then, after the reading of the image stored in the first image processing memory 506 is completed, the images are read from the second image processing memory 507 in the order of addresses and output to the liquid crystal driving unit 105. In this way, since the images formed on the liquid crystal unit 104 are input to the liquid crystal driving unit 105 in the raster scan order, the trapezoidal shape is obtained without performing the process of combining the images stored in the plurality of image processing units. The corrected images can be output to the liquid crystal driving unit 105 in the order of the scanning direction.

  If it is determined in S804 that the image area after the trapezoidal correction process in at least one image processing unit does not include the image to be output to the liquid crystal driving unit 105, the control unit 101 moves the process to S806 and performs the keystone correction. The keystone correction process is completed without performing the process. That is, in the trapezoidal correction process, even if an image included in each image processing unit as an overlap region is referred to, a correction image in a corresponding region among images obtained by dividing the image formed on the liquid crystal unit 104 into four parts in the vertical direction is displayed. If it cannot be generated, the control unit 101 determines not to perform the keystone correction process. That is, the limit amount of the trapezoid correction process in the trapezoid correction unit 502 is limited by the predetermined number of lines in the overlap region.

  In this case, the keystone correction process in the keystone correction unit 502 of each image processing unit is not performed. However, in response to an instruction from the control unit 101, each image processing memory control unit deletes information on an image in an unnecessary area from the address so that only an image not including the overlap area is stored in each image processing memory. Then, the control unit 101 controls reading from each image processing memory so that an image obtained by dividing the projection image into four parts is sequentially output from the image processing unit to the liquid crystal driving unit 105.

  If it is determined that the keystone correction processing is not applied, the control unit 101 stores an OSD image stored in a non-illustrated non-volatile memory, for example, indicating that the correction processing is not performed because the keystone correction limit is exceeded. You may make it project on a screen via the projection optical system 107. In addition, the control unit 101 may output a warning sound from a speaker (not shown) in order to notify the user that the keystone correction process cannot be applied.

  In the present embodiment, even if an image including an overlap region is used, if it is determined that an image obtained by equally dividing an image formed on the liquid crystal unit 104 after the trapezoid correction process cannot be output, the keystone correction process is not performed. However, the implementation of the present invention is not limited to this. For example, the control unit 101 may cause the liquid crystal driving unit 105 to output an image that has been subjected to the maximum amount of trapezoidal correction using the overlap region image in each image processing unit. In the present embodiment, when the first image processing memory, the second image processing memory, the third image processing memory, and the fourth image processing memory are read in order from the head address of the image processing memory included in each image processing unit and projected, It has been described that the projection is performed in the raster scan order. However, the embodiment of the present invention is not limited to this. When projecting with keystone correction, when projecting without performing keystone correction at least when reading and projecting in the same address order as that read when projecting, in raster scan order. It only has to be projected.

  As described above, when projecting a projection image by performing trapezoid correction, the projection apparatus according to the present embodiment divides the projection image, and applies the trapezoid correction to each of the correction images obtained by applying the trapezoid correction. Reading is performed in the same order of addresses as before application, and a plurality of corrected images are projected as one connected image. Specifically, when the projection apparatus divides one projection image into a plurality of divided images, the projection device divides the projection image so as to have an image area overlapping between at least adjacent divided images. The trapezoidal correction of the correction amount is applied to each of the plurality of divided images, and the images extracted from the obtained corrected images are stored in a plurality of storage areas each having an address structure. Then, the images extracted from the plurality of corrected images read in the order of the scanning direction are connected by reading out the images in the same address order as the read-out address order when projecting without performing keystone correction from each of the plurality of storage areas. The projected image is projected as a single image. Note that when storing images in a plurality of storage areas, the projected image is equally divided into the number of divided images, and the trapezoidal distortion of the projected image is corrected when projected onto the projection plane. An image that is connected as an image is extracted from each of the plurality of corrected images and stored.

  As a result, the projection apparatus performs trapezoidal correction on the divided image by a plurality of image processing circuits, and then performs an image on which the trapezoid correction has been performed on the projection image without performing a process of combining the plurality of corrected images. An effect equivalent to reading and projecting in order in the scanning direction can be obtained.

(Embodiment 2)
In the first embodiment, an image obtained by equally dividing the projection image in the vertical direction and an image having a predetermined number of lines from an image adjacent to the equally divided image are output to each image processing unit to perform trapezoid correction processing. The method of extracting the image of the necessary area and outputting it to the liquid crystal drive unit has been described. In the present embodiment, unlike the method described above, the tilt of the projection device is set so that the corrected image obtained by performing the keystone correction process on the image output to each image processing unit becomes the image output to the liquid crystal driving unit as it is. A method of changing the projection image dividing method will be described. Note that a liquid crystal projector, which is an example of the projection apparatus according to the present embodiment described below, has the same functional configuration as that of the first embodiment, and thus description of each block is omitted.

  In this embodiment, the image that is trapezoidally corrected by the trapezoid correction unit 502 of each image processing unit and output to the liquid crystal drive unit 105 is an image that is finally formed on the liquid crystal unit 104 in two orthogonal directions passing through the center. The image is divided into four lines. For example, when the liquid crystal projector 100 projects from below the screen, each image processing is performed so that the liquid crystal unit 104 is formed with the trapezoidal correction and the images as shown in FIGS. The corrected image divided as shown in FIG. Further, when the liquid crystal projector 100 projects from the left side of the screen, each image processing unit as shown in the figure so that an image as shown in FIGS. 11 (f) and 11 (h) is formed on the liquid crystal unit 104. The divided corrected image is output. In this case, the image dividing unit 133 outputs images to be output to the respective image processing units based on the positional relationship between the liquid crystal projector 100 and the screen that is the projection surface, respectively. As shown in g), the division method is output differently. In the example of FIG. 11, it can be seen that the division center, which is the intersection of the lines dividing the projection image, is changed depending on the positional relationship between the liquid crystal projector 100 and the screen. The division center of the projected image may be changed continuously according to the tilt angle of the liquid crystal projector 100, or a predetermined number of division points are prepared in advance, and an optimum division point for the tilt angle amount is selected from them. It may be a system to do. In the present embodiment described below, a method for continuously changing the division center of the projection image according to the former inclination angle amount will be described.

  In the following, processing relating to specific keystone correction performed by the image dividing unit 133 and each image processing unit will be described.

  First, the image dividing unit 133 divides one projection image related to one frame of the input video signal into four regions divided by two orthogonal lines passing through the center of the image as described above, and performs the first division. Store from the memory 201 to the fourth divided memory 204. Specifically, as shown in FIG. 9A, when the projected image is an image of Ht × Vt pixels, each divided memory has four projected images of (Ht / 2) × (Vt / 2) pixels. Images classified into regions are stored.

  The image dividing unit 133 divides the projected image of one frame input from the video input interface by writing the image of the area to be divided into a divided memory in which writing is permitted. Since the projected image input to the image dividing unit 133 is read in the scanning order according to the scanning direction of the raster scan, in this embodiment, the projected image is divided into four regions by using VSYNC and HSYNC.

  The first divided memory control unit 205, the second divided memory control unit 206, the third divided memory control unit 207, and the fourth divided memory control unit 208 generate a pulse in a period of one half of the input VSYNC. Generate WVSYNC. In addition, each divided memory control unit generates WHSYNC that generates a pulse in a half period of the input HSYNC. Then, as shown in the time chart of FIG. 9B, each divided memory control unit transmits the write permission signal, the chip select signal, and the write address information to the divided memory using different WVSYNC and WHSYNC. For example, when the regions to be divided are D1 to D4 as shown in FIG. 9A and each is stored separately from the first divided memory 201 to the fourth divided memory 204, the upper half of the projected image is stored in the first WVSYNC. D1 and D3 are read. When the projected image is read in one line, WHSYNC is generated twice, and in the first WHSYNC, the first divided memory control unit 205 reads the write permission signal so that the pixels in the line D1 are read into the first divided memory 201. Send. In the second WHSYNC, the third divided memory control unit 207 transmits a write permission signal so that the pixels of the line D3 are read into the third divided memory 203. Further, in the next WVSYNC, D2 and D4 in the lower half of the projection image are read, and D2 and D4 are read in two WHSYNC generated by reading one line. That is, in the first WHSYNC, the second divided memory control unit 206 transmits a write permission signal so that the pixels of the line D2 are read into the second divided memory 202. In the second WHSYNC, the fourth divided memory control unit 208 transmits a write permission signal so that the pixels on the line D4 are read into the fourth divided memory 204. In this way, the image signals input in the raster scan order are divided into four divided memories because they are written in the divided memory that is permitted to be written, which is changed by WVSYNC and WHSYNC generated during reading. become.

  The divided image signals stored in the respective divided memories are each image in the subsequent stage according to the projection image division method determined by the positional relationship between the liquid crystal projector 100 and the screen detected by the tilt sensor 134 or the imaging unit 124. Output to the processing unit. In the present embodiment, for example, consider a case where the determined division method divides a projected image by a line shifted from the center in the horizontal direction by the Hs line and in the vertical direction by the Vs line as shown in FIG. That is, the regions Q1, Q2, Q3, and Q4 divided by the division method described above are the first image processing unit 117, the second image processing unit 118, the third image processing unit 119, and the fourth image processing unit 120, respectively. The image is output to

  Here, the operation of each selector when an image divided by the determined division method is output to each image processing unit will be described.

  As shown in FIG. 10A, the output areas Q1 to Q4 each include areas D1 to D4 that are stored in the respective divided memories and are divided into four equal parts as follows.

Area Q1: Vertical direction Vu line + from the beginning of area D1
Area D2 horizontal Hs line from beginning × vertical Vu line from beginning Area Q2: vertical direction Vs line from end of area D1 +
The horizontal direction Hs line from the head of the region D2 × the vertical direction Vs line from the end +
All areas in area D3 +
Area D4 horizontal Hs line from beginning of area D3: Horizontal direction Hu line from end of area D3 × vertical Vu line from start of area D4 Area Q4: Horizontal direction Hu line from end of area D3 × vertical direction Vs line +
A horizontal Hu line from the end of the region D4 In order to output the images (divided images) of the regions Q1 to Q4 divided in this way to each image processing unit, a timing chart representing image data output from each selector As shown in FIG.

  FIG. 12A shows the output from each selector in the horizontal synchronization period, and the image signals stored in the first divided memory 201 to the fourth divided memory 204 are synchronized with the input of the HSYNC pulse. It is read by the memory control unit and output from each selector. Note that the example of FIG. 12A is a timing chart in the case where output from all selectors to each image processing unit is performed simultaneously.

  As shown in the figure, the first selector 209 outputs all the Ht / 2 pixels read from the head address of the first divided memory 201 to the first image processing unit 117, and then starts from the head address of the third divided memory 203. The read Hs pixel is output. In addition, the third selector 211 outputs the Ht / 2−Hs = Hu pixel read from the address offset by Hs pixels from the head address of the third divided memory 203 to the third image processing unit 119. Similarly, the second selector 210 outputs all the Ht / 2 pixels read from the head address of the second divided memory 202 to the second image processing unit 118 and then reads from the head address of the fourth divided memory 204. Output Hs pixels. The fourth selector 212 outputs the Ht / 2−Hs = Hu pixel read from the address offset by Hs pixels from the start address of the fourth divided memory 204 to the fourth image processing unit 120.

  FIG. 12B shows the output from each selector in the vertical synchronization period, and the image signals stored in the first divided memory 201 to the fourth divided memory 204 are synchronized with the input of the VSYNC pulse. It is read by each memory control unit and output from each selector. In the example of FIG. 12B, each selector is a timing chart showing the output of an image output to each image processing unit at the timing of HSYNC.

  As shown in the drawing, the first selector 209 outputs the Vu line read from the head address of the first divided memory 201 to the first image processing unit 117 after HSYNC has passed Vs times. Further, the second selector 210 outputs the Vt / 2−Vu = Vs line read from the address offset by the Vu line from the start address of the first divided memory 201 to the second image processing unit 118. Thereafter, the second selector 210 outputs all Vt / 2 lines read from the head address of the second divided memory 202 to the second image processing unit 118. Similarly, the third selector 211 outputs the Vu line read from the top address of the third divided memory 203 to the third image processing unit 119 after HSYNC has passed Vs times. The fourth selector 212 outputs the Vt / 2−Vu = Vs line read from the address offset by the Vu line from the start address of the third divided memory 203 to the fourth image processing unit 120. Thereafter, the fourth selector 212 outputs all the Vt / 2 lines read from the head address of the fourth divided memory 204 to the fourth image processing unit 120.

  Further, HSYNC and VSYNC are summarized and the output of each selector in the vertical synchronization period is shown in detail as shown in FIG. First, the second selector 210 and the fourth selector 212 first select the images in the regions Q2 and Q4 included in the Vt / 2 line from the vertical Vu line of the image formed in the liquid crystal unit 104 during the first Vs HSYNC. Output. Further, from the time when Vs times HSYNC elapses until Vu times HSYNC elapses, that is, until half the vertical synchronization period, all of the first selector 209 to the fourth selector 212 transfer to each image processing unit. Output image signals. Specifically, the first selector 209 and the third selector 211 output the images of the regions Q1 and Q3 included in the Vu line from the first vertical line of the image formed on the liquid crystal unit 104. The second selector 210 and the fourth selector 212 output the images of the regions Q2 and Q4 included in the Vt / 2 + Vu line from the vertical Vt / 2 line of the image formed on the liquid crystal unit 104. That is, after Vt / 2 times of HSYNC have elapsed, the first selector 209 and the third selector 211 complete the output of the image signal to the first image processing unit 117 and the third image processing unit 119, respectively. Thereafter, the second selector 210 and the fourth selector 212 start from the Vt / 2 + Vu line in the vertical direction of the image formed on the liquid crystal unit 104 until Vs times HSYNC has passed after Vt / 2 times HSYNC has passed. The images of the areas Q2 and Q4 included in are output.

  That is, the reading from each divided memory related to the output from each selector of the image dividing unit 133 of the present embodiment can be completed during Vt / 2 + Vs HSYNCs, and four divisions are performed in one frame period. Image signals of all projected images are read from the memory. In this case, the read clock from each divided memory can be slower than the write clock to the divided memory. However, since the second selector 210 outputs a part of the image signals of the first divided memory 201 and the fourth divided memory 204 to the second image processing unit 118 in addition to all the image signals of the second divided memory 202, it is read out. The clock should be as follows: That is, the read clock from each divided memory must be at least faster than a quarter of the speed of the write clock, and in this way, all image signals of the projected image can be obtained in one frame period. Can be read out to be output to the respective image processing units. By providing each divided memory with a plurality of banks, writing is performed in one bank and reading is performed in the other bank, so that the addresses for writing and reading do not conflict, thereby making memory control easy. .

  In the configuration of the image dividing unit 133 shown in FIG. 2, it is illustrated on the assumption that each divided memory is a two-port memory. However, the present invention is not limited to this, and each divided memory is a one-port memory. There may be. In this case, each divided memory may input a projection image and output to each selector via a data bus.

(Keystone correction processing)
A trapezoid correction process executed by the trapezoid correction unit 502 of this embodiment having such a configuration will be described with reference to the flowchart of FIG. The trapezoid correction process will be described assuming that the control unit 101 starts when the input of the projection image to the image dividing unit 133 is started.

  In step S1301, the control unit 101 causes the divided memory control units of the image dividing unit 133 to equally divide the input projection image and store the divided projection images in the divided memories, so that WVSYNC and WHSYNC as illustrated in FIG. A pulse signal synchronized with is generated. Projected images are input to the image dividing unit 133 in the order of raster scanning, and are written in a divided memory in which pulse signals that can be written from the divided memory control unit are generated.

  In step S <b> 1302, the control unit 101 acquires information on the current tilt angle of the liquid crystal projector 100 from the tilt sensor 134 and grasps the positional relationship between the liquid crystal projector 100 and the screen. From the obtained tilt angle information, the control unit 101 calculates a necessary correction amount so that the image projected on the screen is displayed in a similar shape to the projected image (S1303). Further, the control unit 101 calculates coordinate correspondence information representing coordinates after trapezoid correction processing of all the pixels stored in each divided memory from the correction amount information calculated in S802 (1304).

  In step S1305, the control unit 101 determines a projection image dividing method to be output to each image processing unit, according to the obtained coordinate correspondence information. Specifically, the control unit 101 applies a trapezoidal correction process in each image processing unit using the coordinate correspondence information, and equally divides the image formed on the liquid crystal unit 104 including a line where no pixel value exists. The division method of the projection image is determined so that the image is output. For example, when the projected image is Ht × Vt pixels, the image dividing unit 133 divides the line with the center of the projected image coordinate corresponding to Ht / 2 and vertical coordinate Vt / 2 after the trapezoidal correction process. To do. Therefore, the control unit 101 uses the coordinate correspondence information to calculate the coordinates that are the center of the image after the keystone correction in the projection image.

  In step S1306, the control unit 101 causes each divided memory control unit to read out an image signal from each divided memory as illustrated in FIG. 12 based on the information about the coordinates serving as the image center after the keystone correction calculated in step S1305. , Output to each selector. The control unit 101 also reads the image signal from each divided memory and controls the operation of each selector so that the image signal read from each divided memory is output to a desired image processing unit.

  In step S <b> 1307, the control unit 101 causes all the image processing units to perform the keystone correction process for one frame according to the coordinate correspondence information, and completes the process. In the present embodiment, unlike the above-described first embodiment, the corrected image obtained by performing the keystone correction by the keystone correction unit 502 has the same size as the image obtained by dividing the image formed on the liquid crystal unit 104 into four. Therefore, it is not necessary to extract an image of a necessary area and store it in each image processing memory. In other words, the corrected images output from the trapezoid correcting unit 502 are stored in the order of addresses of the image processing memories.

  The corrected images to which the keystone correction processing is applied and stored in the image processing memory are read out in the order of the addresses of the respective image processing units so that the control unit 101 has the raster scan order of the images formed on the liquid crystal unit 104. Are input to the liquid crystal driver 105. In the present embodiment, since the image including the line having no pixel value formed in the liquid crystal unit 104 is divided into two in the horizontal direction and divided into two in the vertical direction, the image is stored in each image processing memory. The image is read out and output to the liquid crystal drive unit 105 as follows. That is, the control unit 101 reads Ht / 2 pixels from the head address of the first image processing memory 506 for each line up to the Vt / 2 line in the vertical direction of the image formed on the liquid crystal unit 104, and then performs the third operation. Ht / 2 pixels are read from the head address of the image processing memory 508 and output. In the next line, it is assumed that Ht / 2 pixels are read out from the address following the pixels in the already read out line. Further, the control unit 101 starts from the Vt / 2 + 1 line in the vertical direction of the image formed on the liquid crystal unit 104, Ht / 2 pixels from the head address of the second image processing memory 507 for each line, and the fourth image processing memory 509. Ht / 2 pixels are read out from the head address of and output. As in the case of the Vt / 2 line, in the next line, Ht / 2 pixels are read out from the addresses following the pixels in the already read out line. As described above, since the images formed on the liquid crystal unit 104 are input to the liquid crystal drive unit 105 in the raster scan order, the keystone correction is performed without performing the process of combining the images stored in the plurality of image processing units. The images can be output to the liquid crystal drive unit 105 in the order of the scanning direction.

  As described above, when projecting a projection image by performing trapezoidal correction, the projection apparatus according to the present embodiment divides the projection image, reads a correction image to which each of the keystone corrections is applied in order of addresses, The corrected image is projected as a single image. Specifically, when the projection apparatus divides one projection image into a plurality of divided images, each of the correction images obtained by performing the trapezoidal correction is an image obtained when the keystone correction is performed on the projection image. The projected image is divided so as to obtain an equally divided image. A trapezoidal correction with a correction amount is applied to each of the divided divided images, and the obtained corrected images are stored in a plurality of storage areas each having an address structure. Then, by reading out the images in the order of addresses from each of the plurality of storage areas, the images extracted from the plurality of corrected images read in the order of the scanning direction are projected onto the projection plane as a connected image.

  As a result, the projection apparatus performs trapezoidal correction on the divided image by a plurality of image processing circuits, and then performs an image on which the trapezoid correction has been performed on the projection image without performing a process of combining the plurality of corrected images. An effect equivalent to reading and projecting in order in the scanning direction can be obtained.

(Embodiment 3)
In the second embodiment described above, the read clock of each divided memory of the image dividing unit 133 is set to a speed equal to or higher than ¼ of the write clock speed, and all projected images are read out in one frame and sent to each image processing unit. And various correction processes including trapezoidal correction are performed. In the present embodiment, a process when the ratio of the divided image output to the specific image processing unit to the projection image is large will be described. That is, for example, as shown in FIG. 10B, when the image of the region Q2 occupies an area of about 70 to 80% of the projected image, various correction processes of the second image processing unit 118 are processed within the vertical synchronization period. For this purpose, it is necessary to secure a speed of about 70 to 80% of the write clock to each divided memory.

  For this reason, the image dividing unit 133 of this embodiment further includes a PLL 213 (Phase Locked Loop). The PLL 213 is a block that generates a clock for completing image processing in all image processing units within a vertical synchronization period in accordance with a projection image dividing method determined from the positional relationship between the liquid crystal projector 100 and the screen. Specifically, the PLL 213 controls the reading of each divided memory in the image dividing unit 133 by generating a new reading clock rclk that is obtained by multiplying the reading clock of the divided memory. Further, the clock generated by the PLL 213 is input to each image processing unit together with the image signal, and also serves as a clock when performing correction processing in each image processing unit. That is, in order to align the processing operation speed in each image processing unit, the processing speed is not increased only in the image processing unit to which a region having a large ratio to the projected image is input, but synchronization in each image processing unit is easy to take. As described above, all image processing units follow the clock generated by the PLL 213. The PLL 213 receives a write clock wclk to each divided memory, a common clock, and a multiplication number of the read clock determined based on the positional relationship between the liquid crystal projector 100 and the screen, and is used as a projection image division method. An optimal read clock is generated.

  In addition, as shown in FIG. 10A, when the region Q2 having a relatively small inclination angle and the largest area is about 20 to 30% larger than the area obtained by equally dividing the projection image into four parts, the following is considered. be able to. That is, if the speed of the read clock of each divided memory and the operation clock of the second image processing unit 118 is about 20 to 30% faster than a quarter of the speed of the write clock, each image within the vertical synchronization period. The correction process in the processing unit can be completed. That is, when the reading clock from each image division memory and the processing clock in each image processing unit are made faster, the power consumption becomes larger than when operating with a normal clock. It is desirable to use a limited operating clock.

(Keystone correction processing)
A trapezoid correction process executed by the trapezoid correction unit 502 of this embodiment having such a configuration will be described with reference to the flowchart of FIG. The trapezoid correction process will be described assuming that the control unit 101 starts when the input of the projection image to the image dividing unit 133 is started. Further, in the trapezoidal correction process of the present embodiment, the steps for performing the same processes as those of the second embodiment described above are denoted by the same reference numerals, the description thereof is omitted, and only the processes characteristic to the present embodiment are described. .

  When the information of the coordinates serving as the image center after the keystone correction is calculated in S1305, in S1401, the control unit 101 specifies the region of the image having the maximum area among the images into which the projection image is divided. Then, the control unit 101 calculates the multiplication number of the read clock that should be generated in the PLL 213 from the ratio between the area of the region and the area of the projection image, and inputs the multiplication number to the PLL 213. When the multiplication number of the read clock is input by the control unit 101, the PLL 213 generates a read clock from each divided memory and outputs it to each divided memory control unit and each subsequent image processing unit.

  In step S <b> 1306, the control unit 101 receives the image signal from each divided memory to each divided memory control unit using the clock generated by the PLL 213 based on the coordinate information of the projected image that is the center of the image after trapezoid correction calculated in step S <b> 1305. Read and output to each selector. The control unit 101 also reads the image signal from each divided memory and controls the operation of each selector so that the image signal read from each divided memory is output to a desired image processing unit. In step S <b> 1307, the control unit 101 causes all image processing units to perform the keystone correction process for one frame according to the coordinate correspondence information using the clock generated by the PLL 213, and completes the process.

  As described above, when projecting a projection image by performing trapezoidal correction, the projection apparatus according to the present embodiment divides the projection image, reads a correction image to which each of the keystone corrections is applied in order of addresses, The corrected image is projected as a single image. Specifically, when the projection apparatus divides one projection image into a plurality of divided images, each of the correction images obtained by performing the trapezoidal correction is an image obtained when the keystone correction is performed on the projection image. The projected image is divided so as to obtain an equally divided image. A trapezoidal correction with a correction amount is applied to each of the divided divided images, and the obtained corrected images are stored in a plurality of storage areas each having an address structure. Then, by reading out images from each of the plurality of storage areas in the same address order as the order of addresses read when projecting without performing keystone correction, images extracted from the plurality of corrected images read in the order of the scanning direction can be obtained. The projected image is projected onto the projection plane as one connected image.

  As a result, the projection apparatus performs trapezoidal correction on the divided image by a plurality of image processing circuits, and then performs an image on which the trapezoid correction has been performed on the projection image without performing a process of combining the plurality of corrected images. An effect equivalent to reading and projecting in order in the scanning direction can be obtained.

  Further, when dividing the projected image into a plurality of divided images and when performing trapezoidal correction on each divided image, the size of the projected image with respect to the size of the divided image having the largest area among the plurality of divided images. Change the clock speed for processing by a percentage. As a result, it is not necessary to use an unnecessarily high-speed clock for processing, and an increase in power consumption accompanying an increase in clock speed can be suppressed.

(Other embodiments)
The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, or the like) of the system or apparatus reads the program. It is a process to be executed.

Claims (14)

Based on the correction amount of the trapezoid correction process performed on the image projected on the projection surface, the first image including the first region from a predetermined image and a size different from the size of the first region Generating means for generating a second image including a second region;
First correction means for performing a first trapezoid correction process on the first image;
Second correction means for performing a second trapezoid correction process on the second image;
Projection means for projecting an image generated based on the image on which the first trapezoid correction processing has been performed and the image on which the second trapezoid correction processing has been performed on the projection plane ;
A clock provided in common to the first correction unit and the second correction unit, the clock for controlling execution of the first trapezoid correction process and the second trapezoid correction process possess control means for controlling the providing and,
The control means is faster as the target of the keystone correction process is larger, and the clock based on the larger one of the first image and the second image is set to the first correction means and the first correction means. A projection apparatus controlled to be provided to the correction means . The projection apparatus according to claim 1, wherein the image on which the first trapezoid correction process has been performed and the image on which the second trapezoid correction process has been performed are read in a raster scan order. The projection apparatus according to claim 1, wherein the first image includes an image corresponding to a part of a region included in the second image. 4. The projection apparatus according to claim 1, wherein the second image includes an image corresponding to a part of a region included in the first image. 5. The projection apparatus according to claim 1, wherein the correction amount is acquired by using an inclination sensor. The projection apparatus according to claim 1, wherein the correction amount is acquired by using an imaging unit. Having an operation means for receiving an operation from the user;
  The projection apparatus according to claim 1, wherein the correction amount is acquired based on an operation received by the operation unit. Based on the correction amount of the trapezoid correction process performed on the image projected on the projection surface, the first image including the first region from a predetermined image and a size different from the size of the first region Generating a second image including a second region;
Performing a first trapezoidal correction process on the first image;
Performing a second trapezoidal correction process on the second image;
Projecting an image generated based on the image on which the first trapezoid correction processing has been performed and the image on which the second trapezoid correction processing has been performed on the projection plane ;
A clock provided in common for the first correction process and the second correction process, the clock for controlling execution of the first trapezoid correction process and the second trapezoid correction process. possess and controlling, the,
In the controlling step, the larger the target of the keystone correction process, the higher the speed, and the clock based on the larger one of the first image and the second image is the first correction process and the projection method, characterized in Rukoto provided for the second correction processing. 9. The projection method according to claim 8, wherein the image on which the first trapezoid correction process has been performed and the image on which the second trapezoid correction process has been performed are read out in a raster scan order. The projection method according to claim 8, wherein the first image includes an image corresponding to a part of a region included in the second image. The projection method according to claim 8, wherein the second image includes an image corresponding to a part of a region included in the first image. The projection method according to claim 8, wherein the correction amount is acquired by using an inclination sensor. The projection method according to claim 8, wherein the correction amount is acquired by using an imaging unit. The projection method according to claim 8, wherein the correction amount is acquired based on an operation from a user received by an operation unit.

Images