CN106934788B - Rapid extraction method of straight line inclination angle - Google Patents
Rapid extraction method of straight line inclination angle Download PDFInfo
- Publication number
- CN106934788B CN106934788B CN201511018905.2A CN201511018905A CN106934788B CN 106934788 B CN106934788 B CN 106934788B CN 201511018905 A CN201511018905 A CN 201511018905A CN 106934788 B CN106934788 B CN 106934788B
- Authority
- CN
- China
- Prior art keywords
- data
- image
- projection
- coordinates
- taking
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection 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 invention relates to a method for quickly extracting a straight line inclination angle, which comprises the following steps: the image input unit carries out binarization processing on the image to obtain binarization data and coordinates, and sends the binarization data and the coordinates to the plurality of operation units; each single-angle image coordinate transformation and projection extreme value selection unit performs center rotation on the coordinates of the data in the image according to a set angle to obtain rotated coordinates; taking the rotated horizontal coordinate as a read-write address of the dual-port RAM, and taking the binary data as projection accumulated data of the dual-port RAM so as to complete the projection of the rotated image in the horizontal direction; selecting a maximum value in the projection data; the output result selection means selects the maximum value among the output results transmitted from the respective arithmetic means, and the corresponding set angle is used as the inclination angle of the straight line. The invention realizes the single-angle image coordinate transformation and projection extreme value selection unit, only needs to store less projection result data, does not need a large-capacity storage space to buffer the image data, and has high operation efficiency.
Description
Technical Field
The invention relates to the fields of image inclination correction, geometric figure feature extraction and the like in the field of image processing, in particular to a method for quickly extracting a straight line inclination angle.
Background
Line tilt angle extraction is a very important aspect in image processing. The method comprises the steps of firstly realizing image coordinate transformation and projection extreme value selection units, performing projection operation on image data at a given angle, only caching single-line data, realizing multi-path unit parallel operation control by a parallel operation control module, obtaining a middle maximum value by taking eight comparison data as a group, then taking the result of each group as the comparison data again, and finally obtaining a maximum value, thereby obtaining the inclination angle of a straight line.
Disclosure of Invention
The invention is applied to the rapid extraction of the linear inclination angle in image processing, can be widely applied to the field of image processing or detection as a general module, and is beneficial to improving the performance and the speed of a processing system.
The technical scheme of the invention comprises the following steps: a method for rapidly extracting a straight line inclination angle comprises the following steps:
the image input unit carries out binarization processing on the gray level image containing the straight line to obtain binarization data and coordinates of the data in the image and sends the binarization data and the coordinates of the data in the image to a plurality of single-angle image coordinate transformation and projection extreme value selection units;
each single-angle image coordinate transformation and projection extreme value selection unit performs center rotation on the coordinates of the data in the image according to a set angle to obtain rotated coordinates; taking the rotated horizontal coordinate as a read-write address of the dual-port RAM, and taking the binary data as projection accumulated data of the dual-port RAM so as to complete the projection of the rotated image in the horizontal direction; selecting the maximum value in the projection data as an output result and sending the output result to an output result selection unit;
the output result selection means selects the maximum value among the output results transmitted from the respective arithmetic means, and the set angle corresponding to the maximum value is set as the inclination angle of the straight line.
The binarization processing comprises the following steps:
taking the straight line as a foreground, taking 1 during binarization, and taking 0 for a background to obtain a binary image;
and in the output process of the binary image data, row-column counting is adopted to obtain binary data and coordinates of the data in the image.
The method comprises the following steps of taking the rotated horizontal coordinate as a read-write address of a dual-port RAM, taking binary data as projection accumulated data of the dual-port RAM, and completing projection of a rotated image in the horizontal direction, wherein the projection accumulated data comprises the following steps:
reading the written data by taking the rotated horizontal coordinate as an address for accessing the dual-port RAM, accumulating the written data with the current binary data, and writing the data into the dual-port RAM again;
after the coordinates of all the images and the binary data are input, the projection of the rotated images in the horizontal direction is realized.
The output result selection unit selects the maximum value among the output results sent by the operation units, and includes the following steps:
the output result selection unit groups the output results sent by the operation units, extracts the maximum value from each group as a result, and selects the maximum value from the results of each group.
A method for rapidly extracting a straight line inclination angle is realized by an FPGA.
The invention has the following advantages:
1. the invention can be widely applied to relevant systems needing straight line inclination angle extraction as an effective method.
2. The invention realizes the single-angle image coordinate transformation and projection extreme value selection unit, only needs to store less projection result data, does not need a large-capacity storage space to buffer the image data, and has high operation efficiency.
3. The algorithm structure of the invention fully considers the parallel operation characteristic of the programmable logic device, uses a plurality of rotary operation units to realize multi-angle parallel operation, and can obviously improve the processing speed.
4. The invention uses a modular design method to divide the inclination angle extraction operation process into a plurality of reconfigurable units, and can be very easily embedded into a related application system.
Drawings
FIG. 1 is a schematic view of the structure of the present invention.
Detailed Description
The present invention will be described in further detail below with reference to the accompanying drawings.
As shown in fig. 1, the present embodiment is implemented by using an FPGA, and includes an image input unit, a single-angle image coordinate transformation operation unit, a projection extremum selection unit, and an output result selection unit.
The collected gray image data containing straight lines are sequentially input into an image input unit according to the sequence of rows and columns, the image data input unit firstly carries out binarization processing on the gray image data, the selection of a threshold value during the binarization processing is obtained by a large number of experiments according to the characteristics of gray values of the straight lines and a background in practical application, the gray value where the straight lines are located is used as a foreground and is assigned to be 1, and the gray value where the background is located is 0, so that binary image data are obtained; the binary image data completes coordinate generation of the binary image data by using a row-column counter while inputting n single-angle image coordinate transformation and projection extreme value selection units.
When the number of the single-angle image coordinate transformation operation units and the number of the projection extreme value selection units are selected, the practical application conditions are followed, all possible situations are covered by a small angle range, for example, for the application with a small angle, the number of the operation units can be properly selected, so that the consumption of operation resources is reduced, and the operation efficiency of the algorithm is improved. In this embodiment, 30 arithmetic units are used.
Each single-angle image coordinate transformation operation unit and the projection extreme value selection unit firstly rotate the coordinate data sent by the image input module by taking the image center as a rotation center according to a set angle (in the embodiment, the serial number of the operation unit is used as the angle, namely 0-29 degrees), so as to obtain horizontal and vertical coordinates after the image rotation; then, the obtained horizontal coordinate is used as an access address of the azimuth dual-port RAM, written data of the dual-port RAM is read out to be added with image data sent by the image input module, and then the data is written back to the dual-port RAM, namely, the operation process of reading, repairing and rewriting is completed, it needs to be noted that the situation that two adjacent access addresses are the same when the dual-port RAM is accessed can occur, and the solution method is as follows: using a register to cache the last operation result, if the two access addresses are the same, using the value of the register as the result of reading the dual-port RAM, otherwise, selecting the dual-port RAM to read the value, thereby avoiding the occurrence of read-write collision; after the binary data of the whole image is input, the projection of the image data in the horizontal direction is finished; and after the projection data are obtained, the projection data are sequentially output, the maximum value is extracted through a comparison storage operation, and the maximum value is obtained and used as the operation result of the single-angle image coordinate transformation operation unit and the projection extreme value selection unit.
The output result selection unit firstly carries out sequential grouping processing on the data sent by each single angle operation unit, if 0-7 degrees are the 0 th group, 8-15 degrees are the 1 st group, and the like, each group extracts the maximum value in parallel, the eight grouping processing is completed, then the maximum value is selected from the results, and the rotation set angle corresponding to the value is the inclined angle of the image.
Claims (3)
1. A method for rapidly extracting a straight line inclination angle is characterized by comprising the following steps:
the image input unit carries out binarization processing on the image containing the straight line to obtain binarization data and coordinates of the data in the image and sends the binarization data and the coordinates of the data in the image to the multiple single-angle image coordinate transformation and projection extreme value selection units;
each single-angle image coordinate transformation and projection extreme value selection unit performs center rotation on the coordinates of the data in the image according to a set angle to obtain rotated coordinates; taking the rotated horizontal coordinate as a read-write address of the dual-port RAM, and taking the binary data as projection accumulated data of the dual-port RAM so as to complete the projection of the rotated image in the horizontal direction; selecting the maximum value in the projection data as an output result and sending the output result to an output result selection unit;
the method comprises the following steps of taking the rotated horizontal coordinate as a read-write address of a dual-port RAM, taking binary data as projection accumulated data of the dual-port RAM, and completing projection of a rotated image in the horizontal direction, wherein the projection accumulated data comprises the following steps:
reading the written data by taking the rotated horizontal coordinate as an address for accessing the dual-port RAM, accumulating the written data with the current binary data, and writing the data into the dual-port RAM again;
after the coordinates of all the images and the binary data are input, the projection of the rotated images in the horizontal direction is realized;
the output result selection unit selects the maximum value from the output results sent by the operation units, and the set angle corresponding to the maximum value is used as the inclination angle of the straight line;
the output result selection unit selects the maximum value among the output results sent by the operation units, and includes the following steps:
the output result selection unit groups the output results sent by the operation units, extracts the maximum value from each group as a result, and selects the maximum value from the results of each group;
wherein, 0-7 degrees are 0 th group, 8-15 degrees are 1 st group, and so on, each group extracts the maximum value in parallel, the eight groups are completed, then the maximum value is selected from the results, and the rotation setting angle corresponding to the value is the inclination angle of the image.
2. The method for rapidly extracting the straight line inclination angle according to claim 1, wherein the binarization processing comprises the steps of:
taking the straight line as a foreground, taking 1 during binarization, and taking 0 for a background to obtain a binary image;
and in the output process of the binary image data, row-column counting is adopted to obtain binary data and coordinates of the data in the image.
3. The method for rapidly extracting the straight line inclination angle according to claim 1, which is implemented by an FPGA.
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 CN106934788A (en) | 2017-07-07 |
| CN106934788B true 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 (4)
| 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 |
| CN105096298A (en) * | 2014-05-08 | 2015-11-25 | 东北大学 | Grid feature point extraction method based on fast line extraction |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9834153B2 (en) * | 2011-04-25 | 2017-12-05 | Magna Electronics Inc. | Method and system for dynamically calibrating vehicular cameras |
| CN104766078B (en) * | 2015-04-02 | 2018-06-22 | 四川九洲电器集团有限责任公司 | The inclination angle detection method of target image and inclination angle detection device |
-
2015
- 2015-12-30 CN CN201511018905.2A patent/CN106934788B/en active Active
Patent Citations (4)
| 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 |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106934788A (en) | 2017-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106203619B (en) | Data optimized neural network traversal | |
| US20140270364A1 (en) | Performing object detection operations via a graphics processing unit | |
| CN107748723B (en) | Storage method and access device supporting conflict-free stepping block-by-block access | |
| US20110110607A1 (en) | Image processor and method of image rotation | |
| CN104331861B (en) | Image rotating method and system | |
| CN106530209A (en) | FPGA-based image rotation method and apparatus | |
| CN103760525B (en) | Completion type in-place matrix transposition method | |
| Ngo et al. | Resource-aware architecture design and implementation of Hough transform for a real-time iris boundary detection system | |
| CN115035128B (en) | Image overlapping sliding window segmentation method and system based on FPGA | |
| US11106968B1 (en) | Circuit arrangements and methods for traversing input feature maps | |
| CN112396072A (en) | Image classification acceleration method and device based on ASIC and VGG16 | |
| CN105550979A (en) | High-data-throughput texture cache hierarchy structure | |
| CN106934788B (en) | Rapid extraction method of straight line inclination angle | |
| CN110490312B (en) | Pooling calculation method and circuit | |
| CN110930423B (en) | Object edge feature recognition and extraction method | |
| CN110532219B (en) | FPGA-based ping-pong data storage removing method | |
| CN108920097B (en) | Three-dimensional data processing method based on interleaving storage | |
| CN113900813B (en) | Blind pixel filling method, system and device based on double-port RAM | |
| CN108182169A (en) | Efficient FFT implementation methods in a kind of MTD wave filters | |
| CN104333676A (en) | Defective pixel shielding method and system | |
| CN110569684B (en) | Image binarization method for small-capacity cache | |
| CN110264488B (en) | Binary image edge extraction device | |
| CN112837256B (en) | Circuit system and detection method for Harris corner detection | |
| CN105225199A (en) | A kind of method and system of equidistant cylindrical surface projecting in real time | |
| CN106649136B (en) | Data storage method and storage device |
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 |