CN106934788A - A kind of rapid extracting method of linear angle of inclination - Google Patents

A kind of rapid extracting method of linear angle of inclination Download PDF

Info

Publication number
CN106934788A
CN106934788A CN201511018905.2A CN201511018905A CN106934788A CN 106934788 A CN106934788 A CN 106934788A CN 201511018905 A CN201511018905 A CN 201511018905A CN 106934788 A CN106934788 A CN 106934788A
Authority
CN
China
Prior art keywords
data
coordinate
projection
image
postrotational
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.)
Granted
Application number
CN201511018905.2A
Other languages
Chinese (zh)
Other versions
CN106934788B (en
Inventor
岳成海
佟新鑫
朱丹
王玉良
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenyang Institute of Automation of CAS
Original Assignee
Shenyang Institute of Automation of CAS
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenyang Institute of Automation of CAS filed Critical Shenyang Institute of Automation of CAS
Priority to CN201511018905.2A priority Critical patent/CN106934788B/en
Publication of CN106934788A publication Critical patent/CN106934788A/en
Application granted granted Critical
Publication of CN106934788B publication Critical patent/CN106934788B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection

Landscapes

  • Engineering & Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Processing (AREA)

Abstract

The present invention relates to a kind of rapid extracting method of linear angle of inclination, comprise the following steps:Image is carried out binary conversion treatment by image input units, obtains binaryzation data and coordinate is sent to multiple arithmetic elements;Coordinate of each single angular image coordinate transform with projection extreme value select unit by data in the picture carries out central rotation according to set angle, obtains postrotational coordinate;Using the coordinate of postrotational horizontal direction as the read/write address of dual port RAM, binaryzation data complete the projection in the horizontal direction of postrotational image as the projection cumulative data of dual port RAM with this;Maximum is selected in data for projection;Output result select unit selects maximum in the output result that each arithmetic element sends, corresponding set angle as the straight line inclination angle.The present invention realizes single angular image coordinate transform with projection extreme value select unit, it is only necessary to store less projection result data, without massive store spatial cache view data, operation efficiency is high.

Description

A kind of rapid extracting method of linear angle of inclination
Technical field
The present invention relates to the field such as image slant correction, geometric figure feature extraction in image processing field, tool Saying for body is a kind of rapid extracting method of linear angle of inclination.
Background technology
Linear angle of inclination extraction is the highly important aspect in image procossing.Related algorithm is a lot, but difficult To realize quick treatment, or operand is larger, it is necessary to big caching process, proposes a kind of linear angle of inclination Rapid extracting method, image coordinate conversion is realized first is with projection extreme value select unit, by view data Project is made with given angle, only single line of data need to be cached, by multichannel unit concurrent operation control module reality The concurrent operation control of existing multi-angle, compares data and obtains centre as one group and be most worth using eight, then with every The result of group as data are compared, finally gives maximum again, so as to obtain the angle of inclination of straight line, should Processing method drastically increases arithmetic speed.
The content of the invention
The present invention is applied to the rapid extraction at image procossing cathetus angle of inclination, can be as general module quilt It is widely used in image procossing or detection field, is conducive to improving the performance and speed of processing system.
Technical scheme includes:A kind of rapid extracting method of linear angle of inclination, comprises the following steps:
Gray level image containing straight line is carried out binary conversion treatment by image input units, obtain binaryzation data with And each data coordinate in the picture sends single with projection extreme value selection to the angular image coordinate transform of multiple lists Unit;
Each single angular image coordinate transform and projection extreme value select unit by data coordinate in the picture according to Set angle carries out central rotation, obtains postrotational coordinate;Using the coordinate of postrotational horizontal direction as The read/write address of dual port RAM, binaryzation data complete rotation as the projection cumulative data of dual port RAM with this Image projection in the horizontal direction after turning;In data for projection select maximum as output result send to Output result select unit;
Output result select unit selects maximum in the output result that each arithmetic element sends, the maximum Corresponding set angle as the straight line inclination angle.
The binary conversion treatment is comprised the following steps:
Using straight line as prospect, 1 is taken during binaryzation, background takes 0, obtains bianry image;
Binary image data obtains binaryzation data and each data in figure in output procedure using line count Coordinate as in.
The coordinate using postrotational horizontal direction as dual port RAM read/write address, binaryzation data make Be the projection cumulative data of dual port RAM, with this complete the projection in the horizontal direction of postrotational image include with Lower step:
Using the coordinate of postrotational horizontal direction as the address for accessing dual port RAM, data have been write in reading, with After current binaryzation data add up, dual port RAM is write again;
After the completion of the coordinate and binaryzation data input of all images, postrotational image is realized in level side To projection.
The output result select unit selected in the output result that each arithmetic element sends maximum include with Lower step:
Be grouped for the output result that each arithmetic element sends by output result select unit, and each group is extracted maximum Value from the result of each group as a result, choose maximum again.
A kind of rapid extracting method of linear angle of inclination is realized by FPGA.
Advantage of the present invention is as follows:
1. the present invention can be widely used in the correlation for needing straight line degree of tilt to extract as a kind of effective ways In system.
2. the present invention realize single angular image coordinate transform with projection extreme value select unit, it is only necessary to storage compared with Few projection result data, without massive store spatial cache view data, operation efficiency is high.
3. algorithm structure of the invention fully takes into account the concurrent operation characteristic of PLD, using many Individual twiddle operation unit, realizes the concurrent operation of multi-angle, can significantly improve processing speed.
4. the present invention uses modular design method, and inclination angle extraction calculating process is divided into multiple restructural lists Unit, very can readily be embedded into related application system.
Brief description of the drawings
Fig. 1 is schematic structural view of the invention.
Specific embodiment
The present invention is described in further detail below by with reference to accompanying drawing.
As shown in figure 1, the present embodiment is realized using FPGA, including image input units, single angular image are sat Mark transform operation unit and projection extreme value select unit, output result select unit.
The greyscale image data containing straight line for collecting sequentially inputs image input units according to ranks order, Greyscale image data is carried out binary conversion treatment by view data input block first, threshold value during binary conversion treatment Selection is obtained with the characteristic of the gray value of background according to practical application cathetus by many experiments, where straight line Gray value is entered as 1 as prospect, and the gray value where background takes 0, so as to obtain binary image data; Binary image data makes while n single angular image coordinate transform is input into projection extreme value select unit The Coordinate generation of binary image data is completed with line count device.
In number of the single angular image coordinate transform arithmetic element of selection with projection extreme value select unit, it then follows Practical application condition, all possible situation is covered with less angular range, such as less for angle In, less arithmetic element number can be suitably chosen, the consumption of calculation resources is reduced with this, improved The operation efficiency of algorithm.30 arithmetic elements are used in the present embodiment.
Image is input into mould by each single angular image coordinate transform arithmetic element first with projection extreme value select unit The coordinate data that block sends according to setting angle (the present embodiment using arithmetic element sequence number as angle, That is 0-29 degree) do rotation processing by pivot of picture centre, obtain the level after image rotation with it is vertical Coordinate, due to decimal occurs during twiddle operation, it is therefore desirable to to coordinate do quantification treatment, Completed using the strategy for rounding up;Then using the horizontal coordinate that obtains as the access of orientation dual port RAM Location, first by the view data for having write data read-out to be sent with image input module of dual port RAM do plus and Computing, is then written back in dual port RAM again, that is, complete to read the calculating process that modification is write, it should be noted that The adjacent identical situation of reference address twice is likely to occur when accessing dual port RAM, solution is:Use one The operation result of individual register cache last time, with the value of the register if reference address is identical so twice As dual port RAM result is read, dual port RAM readout is otherwise selected, so as to avoid the generation of read/write conflict; After the completion of the two-value data input of entire image, view data projection in the horizontal direction is also just completed; After obtaining data for projection, it is sequentially output by data for projection, comparing storage computing by one completes maximum Extraction, obtain after maximum as single angular image coordinate transform arithmetic element and projection extreme value select unit Operation result.
Output result select unit is first done at sequential packet the data that each single angle arithmetic element sends Reason, such as 0~7 degree is the 0th group, and 8~15 degree is the 1st group, the like, each group extracts maximum parallel, Eight packet transactions are more than completed, then maximum is chosen from these results again, the rotation corresponding to the value Set angle is exactly the inclination angle of image.

Claims (5)

1. a kind of rapid extracting method of linear angle of inclination, it is characterised in that comprise the following steps:
Image containing straight line is carried out binary conversion treatment by image input units, obtains binaryzation data and each Data coordinate in the picture is sent to the angular image coordinate transform of multiple lists and projection extreme value select unit;
Each single angular image coordinate transform and projection extreme value select unit by data coordinate in the picture according to Set angle carries out central rotation, obtains postrotational coordinate;Using the coordinate of postrotational horizontal direction as The read/write address of dual port RAM, binaryzation data complete rotation as the projection cumulative data of dual port RAM with this Image projection in the horizontal direction after turning;In data for projection select maximum as output result send to Output result select unit;
Output result select unit selects maximum in the output result that each arithmetic element sends, the maximum Corresponding set angle as the straight line inclination angle.
2. the rapid extracting method of a kind of linear angle of inclination according to claim 1, it is characterised in that described two Value treatment is comprised the following steps:
Using straight line as prospect, 1 is taken during binaryzation, background takes 0, obtains bianry image;
Binary image data obtains binaryzation data and each data in figure in output procedure using line count Coordinate as in.
3. the rapid extracting method of a kind of linear angle of inclination according to claim 1, it is characterised in that it is described with The coordinate of postrotational horizontal direction as dual port RAM read/write address, binaryzation data are used as dual port RAM Projection cumulative data, the projection in the horizontal direction of postrotational image is completed with this and is comprised the following steps:
Using the coordinate of postrotational horizontal direction as the address for accessing dual port RAM, data have been write in reading, with After current binaryzation data add up, dual port RAM is write again;
After the completion of the coordinate and binaryzation data input of all images, postrotational image is realized in level side To projection.
4. the rapid extracting method of a kind of linear angle of inclination according to claim 1, it is characterised in that described defeated Go out result select unit selects maximum to comprise the following steps in the output result that each arithmetic element sends:
Be grouped for the output result that each arithmetic element sends by output result select unit, and each group is extracted maximum Value from the result of each group as a result, choose maximum again.
5. the rapid extracting method of a kind of linear angle of inclination according to claim 1, it is characterised in that by FPGA Realize.
CN201511018905.2A 2015-12-30 2015-12-30 Rapid extraction method of straight line inclination angle Active CN106934788B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201511018905.2A CN106934788B (en) 2015-12-30 2015-12-30 Rapid extraction method of straight line inclination angle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201511018905.2A CN106934788B (en) 2015-12-30 2015-12-30 Rapid extraction method of straight line inclination angle

Publications (2)

Publication Number Publication Date
CN106934788A true CN106934788A (en) 2017-07-07
CN106934788B CN106934788B (en) 2020-11-24

Family

ID=59442309

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201511018905.2A Active CN106934788B (en) 2015-12-30 2015-12-30 Rapid extraction method of straight line inclination angle

Country Status (1)

Country Link
CN (1) CN106934788B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102289810A (en) * 2011-08-05 2011-12-21 上海交通大学 Quick rectangle detection method of images high resolution and high order of magnitude
CN102385750A (en) * 2011-06-22 2012-03-21 清华大学 Line matching method and line matching system on basis of geometrical relationship
JP4950290B2 (en) * 2007-06-27 2012-06-13 パナソニック株式会社 Imaging apparatus, method, system integrated circuit, and program
US20150049193A1 (en) * 2011-04-25 2015-02-19 Magna International Inc. Method and system for dynamically calibrating vehicular cameras
CN104766078A (en) * 2015-04-02 2015-07-08 四川九洲电器集团有限责任公司 Slant angle detecting method and slant angle detecting device of target image
CN105096298A (en) * 2014-05-08 2015-11-25 东北大学 Grid feature point extraction method based on fast line extraction

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4950290B2 (en) * 2007-06-27 2012-06-13 パナソニック株式会社 Imaging apparatus, method, system integrated circuit, and program
US20150049193A1 (en) * 2011-04-25 2015-02-19 Magna International Inc. Method and system for dynamically calibrating vehicular cameras
CN102385750A (en) * 2011-06-22 2012-03-21 清华大学 Line matching method and line matching system on basis of geometrical relationship
CN102289810A (en) * 2011-08-05 2011-12-21 上海交通大学 Quick rectangle detection method of images high resolution and high order of magnitude
CN105096298A (en) * 2014-05-08 2015-11-25 东北大学 Grid feature point extraction method based on fast line extraction
CN104766078A (en) * 2015-04-02 2015-07-08 四川九洲电器集团有限责任公司 Slant angle detecting method and slant angle detecting device of target image

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
XIN ZHOU等: "An Efficient Implementation of the Hough Transform Using DSP Slices and Block RAMs on the FPGA", 《2013 IEEE 7TH INTERNATIONAL SYMPOSIUM ON EMBEDDED MULTICORE SOCS》 *
商尔科等: "基于 FPGA 的实时 Hough 变换", 《计算机工程》 *
徐兴东等: "一种检测图像直线倾斜角度的新算法", 《中南民族大学学报( 自然科学版)》 *

Also Published As

Publication number Publication date
CN106934788B (en) 2020-11-24

Similar Documents

Publication Publication Date Title
CN103020890B (en) Based on the visual processing apparatus of multi-level parallel processing
CN103336758B (en) The sparse matrix storage means of a kind of employing with the sparse row of compression of local information and the SpMV implementation method based on the method
CN106021182B (en) A kind of row transposition architecture design method based on Two-dimensional FFT processor
CN108716916B (en) Distributed parallel star point centroid extraction method based on superblock and FPGA implementation device
WO2020125062A1 (en) Image fusion method and related device
Ngo et al. Resource-aware architecture design and implementation of Hough transform for a real-time iris boundary detection system
WO2019076109A1 (en) Method and device for pooling image information, storage medium and processor
CN109740758A (en) A kind of kernel method based on quantum calculation
Fan et al. Implementation of high performance hardware architecture of OpenSURF algorithm on FPGA
Park et al. A vision processor with a unified interest-point detection and matching hardware for accelerating a stereo-matching algorithm
CN106934788A (en) A kind of rapid extracting method of linear angle of inclination
CN102663051B (en) Method and system for searching content addressable memory
CN105260739B (en) Image matching method towards binary features and its system
CN103577160A (en) Characteristic extraction parallel-processing method for big data
CN106097263A (en) Image reconstructing method based on full variation norm image block gradient calculation
CN105069780A (en) SAR image water area detection realization method based on SoPC
CN106775596B (en) A kind of infrared image linear interpolation expansion hardware processing method
CN110473258B (en) Monocular SLAM system initialization algorithm based on point-line unified framework
Zhang et al. Implementation of high performance hardware architecture of face recognition algorithm based on local binary pattern on FPGA
CN114372012A (en) Universal and configurable single-row output system and method for high-energy-efficiency pooling calculation
CN103927269A (en) Reconfigurable configuration information cache system based on block matching and compression method
CN109979565B (en) Dosage calculation optimization method and device based on CUDA
Jia et al. SRI-SURF: A better SURF powered by scaled-RAM interpolator on FPGA
CN110134678A (en) A kind of indexing means of biological data, system and electronic equipment
Xie et al. A real-time FPGA-based architecture of improved ORB

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant