CN2736865Y - Zoom processing system for digital image - Google Patents
Zoom processing system for digital image Download PDFInfo
- Publication number
- CN2736865Y CN2736865Y CN 200420077711 CN200420077711U CN2736865Y CN 2736865 Y CN2736865 Y CN 2736865Y CN 200420077711 CN200420077711 CN 200420077711 CN 200420077711 U CN200420077711 U CN 200420077711U CN 2736865 Y CN2736865 Y CN 2736865Y
- Authority
- CN
- China
- Prior art keywords
- image
- module
- zoom
- register
- host computer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
- 238000012545 processing Methods 0.000 title claims abstract description 17
- 239000000872 buffer Substances 0.000 claims abstract description 30
- 238000011045 prefiltration Methods 0.000 claims abstract description 19
- 238000001914 filtration Methods 0.000 claims description 30
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000012805 post-processing Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000000034 method Methods 0.000 description 9
- 230000008901 benefit Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 230000008676 import Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Images
Landscapes
- Image Processing (AREA)
Abstract
The utility model relates to the field of image processing, in particular to an image zoom processing system used for digital images. The image zoom processing system of the utility model comprises a pre-filter module, an image lateral dimension zoom module, an image zoom row buffer module, a command sending module, an image zoom coefficient computation module, and an image longitudinal dimension zoom module. The image zoom processing system can enhance the performance of realizing a chip to support images and video.
Description
Technical field
The utility model relates to image processing field, specifically, relates to a kind of pantograph treatment system that is used for digital picture.
Background technology
Along with developing rapidly of multimedia field application products such as computing machine, mobile phone, people are more and more higher to the requirement that image shows, because different display techniques has different requirements to size of images, need change flexibly size of images.For example, the resolution of computing machine has VGA (640 * 480), SVGA (800 * 600), and XVGA (1600 * 1200), but concerning mobile phone, the resolution of display screen is less than QVGA (320 * 240) usually.Therefore, how to realize that image changes flexibly and keeps higher picture quality to become a urgent day by day problem in different resolution.
In the image processing field of computing machine, image zoom realizes by software, typically uses the secondary linear interpolation, cubic curve interpolation scheduling algorithm is handled.Software processes need take very many cpu resources, and is very high to the requirement of Computer Processing device.But for mobile phone, the function of CPU is powerful inadequately, is nowhere near concerning the convergent-divergent of handling video, therefore realizes that by software video scaling is impossible, and this just need solve this problem with hardware.
In general video card, video frequency collection card, use the secondary linear interpolation to realize the image zoom of hardware usually.This algorithm can bring the better image quality, but needs to take more hardware resource (for example multiplier and divider etc.).
The utility model content
The purpose of this utility model is for a kind of image zoom disposal system being provided, improving the performance that chip supports that image and video are realized.
For achieving the above object, the utility model provides scheme as follows:
A kind of pantograph treatment system of digital picture, the view data that instruction that its reception host computer 1 sends and image pretreatment module 2 are sent, according to the instruction that host computer 1 sends view data is carried out the convergent-divergent processing, and give post processing of image module 4 with the image that convergent-divergent is handled, compress or directly show by 4 pairs of images of post processing of image module.
Image zoom disposal system described in the utility model comprises prefilter module 4, image lateral dimension Zoom module 5, image zoom line buffer module 6, order sending module 7, image zoom coefficients calculation block 8 and image longitudinal size Zoom module 9; Pass through order sending module 7 by host computer 1, this image zoom disposal system is activated and makes it to be in normal duty, the image that this moment, prefilter module 4 was brought the image pretreatment module carries out low-pass filtering, the coefficient of this low-pass filter is configured according to the image zoom ratio by host computer 1, sends to prefilter module 4 by order sending module 7; Image zoom coefficients calculation block 8 sends to the order of order sending module 7 according to host computer 1, and the coefficient that calculates convergent-divergent also gives image lateral dimension Zoom module 5 and image longitudinal size Zoom module 9 uses; Image lateral dimension Zoom module 5 is laterally carrying out the convergent-divergent processing according to instruction and the zoom factor that order sending module 7 sends to the image through pre-filtering, and processed images is stored in the image zoom line buffer module 6, image longitudinal size Zoom module 9 carries out the conversion of vertical scale to the image in the image zoom line buffer module 6 more then.
Register module 73 described in the utility model comprises input picture width register 731, input picture height register 732, output image width register 733, output image height register 734, prefilter coefficient register 735, image enlargement factor register 736 and image minification register 737; Before starting image zoom, configure the value of these registers by host computer 1, to guarantee the operate as normal of described system.
Compared with prior art, advantage of the present utility model is: in the image zoom disposal system described in the utility model, do not need bigger image buffer, only need several line buffers, saved the use of storer so greatly; Secondly,, import several capable view data and just can start zoom operations, so postpone minimumly, several row are only arranged, need the delay of a frame unlike classic method because this method is true-time operation; The 3rd, because this method is converted into two one-dimensional filterings with two-dimensional filtering,, significantly reduced the use of multiplier with two one bit image convergent-divergents of convergent-divergent conversion dimension of two bit images, the cost of chip is reduced greatly; At last, because this method uses look-up table to replace divider to come the coefficient of computed image scaler, chip cost is further reduced, and let us know that divider is difficult for realizing.
The utility model purpose, characteristics and advantage are further described with reference to the accompanying drawings in conjunction with the embodiments.
Description of drawings
Fig. 1 is a structured flowchart of the present utility model.
Fig. 2 is the structured flowchart of prefilter module described in the utility model.
Fig. 3 is the structured flowchart of order sending module described in the utility model.
Fig. 4 is the structured flowchart of register module described in the utility model.
Fig. 5 is a reference configuration of register described in the utility model.
Embodiment
With reference to accompanying drawing 1, be structured flowchart of the present utility model.
A kind of pantograph treatment system of digital picture, the view data that instruction that its reception host computer sends and image pretreatment module are sent, according to the instruction that host computer sends view data is carried out the convergent-divergent processing, and give the post processing of image module with the image that convergent-divergent is handled, image is compressed or directly show by the post processing of image module.
Image zoom disposal system described in the utility model comprises prefilter module, image lateral dimension Zoom module, image zoom line buffer module, order sending module, image zoom coefficients calculation block and image longitudinal size Zoom module; Pass through the order sending module by host computer, this image zoom disposal system is activated and makes it to be in normal duty, the image that this moment, the prefilter module was brought the image pretreatment module carries out low-pass filtering, the coefficient of this low-pass filter is configured according to the image zoom ratio by host computer, sends to the prefilter module by the order sending module; The image zoom coefficients calculation block sends to the order of order sending module according to host computer, and the coefficient that calculates convergent-divergent also gives image lateral dimension Zoom module and image longitudinal size Zoom module uses; Image lateral dimension Zoom module is laterally carrying out the convergent-divergent processing according to instruction and the zoom factor that the order sending module sends to the image through pre-filtering, and processed images is stored in the image zoom line buffer module, image longitudinal size Zoom module carries out the conversion of vertical scale to the image in the image zoom line buffer module more then.
With reference to accompanying drawing 2, be the structured flowchart of prefilter module described in the utility model.Prefilter module described in the utility model comprises the transversal filter module, filter rows buffer module and longitudinal filter module; The transversal filter module is at first carried out filtering to image on breadth wise dimension, and horizontal filtered image storage in the filter rows buffer module; Because described system takes seven rank filtering, the filter rows buffer module has stored the view data of 6 row; The longitudinal filter module is taken out image from the filter rows buffer module, and it is carried out vertical filtering.
With reference to accompanying drawing 3, be the structured flowchart of order sending module described in the utility model.Order sending module described in the utility model comprises register address decoding module, register read writing module, register module, and the register address decoding module is deciphered the register address that host computer sends over, and finds the register cell that will operate; The register read writing module is then finished the read-write operation to this register; Register module has then comprised the control register of One's name is legion.
With reference to accompanying drawing 4 are structured flowcharts of register module described in the utility model.Register module described in the utility model comprises input picture width register (IMG_WIDTH), input picture height register (IMG_HEIGHT), output image width register (OUT_WIDTH), output image height register (OUT_HEIGHT), prefilter coefficient register (FIR_COEF), image enlargement factor register (UP_FACTOR) and image minification register (DOWN_FACTOR); Before starting image zoom, configure the value of these registers by host computer, to guarantee the operate as normal of described system.
Before image zoom was handled, host computer at first will be finished the configuration of related register, can operate as normal to guarantee this system.For example, if this system's input picture size is 640 * 480, the output image size is 352 * 288, and the coefficient of wave filter takes 0,1/16,3/16,8/16,3/16,1/16,0, and then the configuration of these registers can be with reference to accompanying drawing 5.When image when the image pretreatment module arrives, these relevant configurations are finished.
When image is imported, at first to carry out filtering to it by prefilter.This system takes 7 * 7 two-dimentional low-pass filtering, and choosing of filter coefficient will be determined according to the ratio of image zoom.After the filtering, the high frequency noise of image is removed, but the size of image still is 640 * 480 (are example with above-mentioned reference configuration), at this moment at first will carry out the convergent-divergent of breadth wise dimension; Because this system takes the bilinearity difference, the difference coefficient need be calculated by the image zoom coefficients calculation block.After finishing horizontal convergent-divergent, the size of every row becomes 352, but height all still is 480, and image storage stores the color numerical value of three images in image zoom line buffer module in this module.And then it is carried out convergent-divergent longitudinally by vertical image zoom module.Picture size just has 640 * 480 to become 352 * 288 like this.
With reference to accompanying drawing 2, for the preposition low-pass filtering of image, this system takes 7 * 7 linear difference, and the color numerical value of a promptly filtered pixel is on average obtained through linear weighted functions by 7 * 7 picture elements on every side, so just 49 multipliers of needs.According to aforementioned, this method is converted into two one-dimensional operations with two-dimentional computing, and takes the filter coefficient of symmetry, for each one-dimensional filtering device (laterally filtering and vertically filtering), only needs 4 multipliers like this.Laterally to be filtered into example, suppose that the pixel that will calculate is Pn ', the Pn-3 before it and the filtering, Pn-2, Pn-1, Pn, Pn+1, Pn+2 is relevant with Pn+3, and then the computing formula of Pn ' is:
Pn’＝(Pn-3+Pn+3)*FIR_COEF0+
(Pn-2+Pn+2)*FIR_COEF1+
(Pn-1+Pn+1)*FIR_COEF2+
Pn*FIR_COEF3
The coefficient of above-mentioned formula median filter is disposed by host computer.This coefficient is relevant with the ratio of image zoom, and it should guarantee that image should not have high frequency noise after convergent-divergent, should have than the distinct image quality simultaneously.
After finishing horizontal filtering, image exists in the filter rows buffer module.The filter rows buffer module has comprised six line buffers, can store the data of six row images.Need seven images of going for the vertical filtering in seven rank, have six in the impact damper, what the 7th row was then real-time imports; Six impact dampers not only offer longitudinal filter after the data input of the 7th row and use, and can override the shared impact damper of original first line data, so just can satisfy use.
For vertical filtering, the logical horizontal filtering of its algorithm is just the same, and unique difference is the difference of calculated direction.
Finish after the filtering, will carry out the convergent-divergent of breadth wise dimension image.Scaling is by register UP_FACTOR and DOWN_FACTOR decision, and such as by 640 * 480 to 352 * 288, its scaling is 11/20, UP_FACTOR=11 then, DOWN_FACTOR=20.Image zoom related to for two steps, at first according to UP_FACTOR the lateral dimension of image was become IMG_WIDTH*UP_FACTOR, and promptly 7040; And then at pixel as output image of per 20 extractions in these 7040 pixel the insides, the output image width is exactly 352 like this.But, in specific implementation, can not store the value of 7040 pixels, here because most all of no use and will lose; We only need to store those and appear at the value of final output image pixel.In the horizontal Zoom method of native system, two counters are arranged, one of them is that it can increase by 1 automatically every pixel in order to the number (PIXEL_CNT) of delegation's pixel in the statistics input picture; Another counter (UPSIZE_COUNT) needs the position of the pixel of extraction in order to calculate every row, when new two field picture arrival or the end of this two field picture delegation, and this counter O reset; Later on everyly come the value of new this counter of pixel just to add UP_FACTOR.UPSIZE_COUNT＞=DOWN_FACTOR is found in the back if new pixel arrives, and an output pixel is arranged in output image, subtracts from UPSIZE_COUNT simultaneously and gets DOWN_FACTOR.Because new pixel must appear at the centre of two input pixels that just arrived, it is just obtained by the weighted average calculation of these two pixels.If two pixels that just arrived are P
1And P
2, then new pixel color value can calculate by following formula:
P
new＝(1-a)*P
1+a*P
2
Wherein coefficient a can be according to new pixel P
NewAnd P
1Distance (being UPSIZE_COUNT) obtain:
a＝UPSIZE_COUNT/UP_FACTOR
Because UP_FACTOR is one six bit,, in system, exist a table to store 64 1/UP_FACTOR for avoiding divider.
Pixel is handled according to same method so one by one, just can obtain the delegation's image through horizontal convergent-divergent, and processed images is stored in the image zoom line buffer module 4.Article three, the image zoom line buffer can store the horizontal convergent-divergent processed images of triplex row, and its working method is such: have two row view data and just can carry out vertical convergent-divergent processing in impact damper; But, therefore need the 3rd impact damper to store the image of input because input picture also might continue to come in when doing vertical processing.Vertically the disposal route of image zoom is with laterally about the same.Can obtain a complete image like this through the convergent-divergent processing.
The pantograph treatment system of a kind of digital picture described in the utility model, be not restricted to listed utilization in instructions and the embodiment, it can be applied to the field of various suitable the utility model fully, for those skilled in the art, can easily realize additional advantage and make amendment, therefore under the situation of the spirit and scope of the universal that does not deviate from claim and equivalency range and limited, the examples shown that the utility model is not limited to specific details, representational equipment and illustrates here and describe.
Claims (4)
1, a kind of pantograph treatment system of digital picture, the view data that instruction that its reception host computer (1) sends and image pretreatment module (2) are sent, according to the instruction that host computer (1) sends view data is carried out the convergent-divergent processing, and give post processing of image module (4) with the image that convergent-divergent is handled, image is compressed or directly show by post processing of image module (4);
It is characterized in that: described image zoom disposal system comprises prefilter module (4), image lateral dimension Zoom module (5), image zoom line buffer module (6), order sending module (7), image zoom coefficients calculation block (8) and image longitudinal size Zoom module (9); Pass through order sending module (7) by host computer (1), this image zoom disposal system is activated and makes it to be in normal duty, the image that prefilter module this moment (4) is brought the image pretreatment module carries out low-pass filtering, the coefficient of this low-pass filter is configured according to the image zoom ratio by host computer (1), sends to prefilter module (4) by order sending module (7); Image zoom coefficients calculation block (8) sends to the order of order sending module (7) according to host computer (1), calculates the coefficient of convergent-divergent and also gives image lateral dimension Zoom module (5) and image longitudinal size Zoom module (9) use; Image lateral dimension Zoom module (5) is laterally carrying out the convergent-divergent processing according to instruction and the zoom factor that order sending module (7) sends to the image through pre-filtering, and processed images is stored in the image zoom line buffer module (6), image longitudinal size Zoom module (9) carries out the conversion of vertical scale to the image in the image zoom line buffer module (6) more then.
2, image zoom disposal system according to claim 1 is characterized in that: described prefilter module (4) comprises transversal filter module (41), filter rows buffer module (42) and longitudinal filter module (43); Transversal filter module (41) is at first carried out filtering to image on breadth wise dimension, and horizontal filtered image storage in filter rows buffer module (42); Because described system takes seven rank filtering, filter rows buffer module (42) has stored the view data of 6 row; Longitudinal filter module (43) is taken out image from filter rows buffer module (42), and it is carried out vertical filtering.
3, image zoom disposal system according to claim 1, it is characterized in that: described order sending module (7) comprises register address decoding module (71), register read writing module (72), register module (73), register address decoding module (71) is deciphered the register address that host computer (1) sends over, and finds the register cell that will operate; Register read writing module (72) is then finished the read-write operation to this register; Register module (73) has then comprised the control register of One's name is legion.
4, image zoom disposal system according to claim 3, it is characterized in that: described register module (73) comprises input picture width register (731), input picture height register (732), output image width register (733), output image height register (734), prefilter coefficient register (735), image enlargement factor register (736) and image minification register (737); Before starting image zoom, configure the value of these registers by host computer (1), to guarantee the operate as normal of described system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200420077711 CN2736865Y (en) | 2004-07-08 | 2004-07-08 | Zoom processing system for digital image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200420077711 CN2736865Y (en) | 2004-07-08 | 2004-07-08 | Zoom processing system for digital image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN2736865Y true CN2736865Y (en) | 2005-10-26 |
Family
ID=35307595
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200420077711 Expired - Lifetime CN2736865Y (en) | 2004-07-08 | 2004-07-08 | Zoom processing system for digital image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN2736865Y (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1293514C (en) * | 2004-07-08 | 2007-01-03 | 北京中星微电子有限公司 | Pantograph treatment system of digital picture |
| WO2010105408A1 (en) * | 2009-03-17 | 2010-09-23 | 华为技术有限公司 | Method and device for image resizing |
| CN101777175B (en) * | 2008-07-31 | 2016-03-23 | 奥多比公司 | Reducing and expanding based on gap is carried out to the image with color-weighted priority |
-
2004
- 2004-07-08 CN CN 200420077711 patent/CN2736865Y/en not_active Expired - Lifetime
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1293514C (en) * | 2004-07-08 | 2007-01-03 | 北京中星微电子有限公司 | Pantograph treatment system of digital picture |
| CN101777175B (en) * | 2008-07-31 | 2016-03-23 | 奥多比公司 | Reducing and expanding based on gap is carried out to the image with color-weighted priority |
| WO2010105408A1 (en) * | 2009-03-17 | 2010-09-23 | 华为技术有限公司 | Method and device for image resizing |
| CN102246223A (en) * | 2009-03-17 | 2011-11-16 | 华为技术有限公司 | Method and device for image resizing |
| CN102246223B (en) * | 2009-03-17 | 2013-09-11 | 华为技术有限公司 | Method and device for image resizing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8180165B2 (en) | Accelerated screen codec | |
| US8837854B2 (en) | Image processing method for boundary resolution enhancement | |
| US7720311B1 (en) | Memory and compute efficient block-based two-dimensional sample-rate converter for image/video applications | |
| CN100356405C (en) | Method and apparatus for changing digital image size | |
| CN1925544A (en) | Image data processing device | |
| CN1957371A (en) | Method and system for viewing and enhancing images | |
| CN110428382B (en) | Efficient video enhancement method and device for mobile terminal and storage medium | |
| CN1761312A (en) | Method and circuit for zooming digital video image based on dual quadratic interpolation | |
| CN1668108A (en) | Resolution changing method and apparatus | |
| Bártovský et al. | Parallel implementation of sequential morphological filters | |
| US20050248589A1 (en) | System and method for rapidly scaling and filtering video data | |
| WO2007014365A2 (en) | Image processing using saltating samples | |
| CN1293514C (en) | Pantograph treatment system of digital picture | |
| CN111145074B (en) | Full liquid crystal instrument image rendering method | |
| CN2736865Y (en) | Zoom processing system for digital image | |
| CN1317884C (en) | A method for realizing integral multiple amplification of image | |
| US8873884B2 (en) | Method and system for resizing an image | |
| CN1835545A (en) | Image scaling device using a single line memory and a scaling method thereof | |
| CN105160622B (en) | The implementation method of image super-resolution based on FPGA | |
| US20090087119A1 (en) | Method and apparatus for arbitrary ratio image reduction | |
| CN106780291B (en) | Real-time distortion image processing accelerating device | |
| CN1175378C (en) | Deivce and method for processing covered picture to become transparent one | |
| CN100379249C (en) | A method and apparatus for changing digital image size during process of shooting | |
| CN100345437C (en) | Multiplierless pyramid filter | |
| JP2004354593A5 (en) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| AV01 | Patent right actively abandoned |
Effective date of abandoning: 20040708 |
|
| C25 | Abandonment of patent right or utility model to avoid double patenting |