CN104598910A - Smart television station caption identification method and system based on gradient direction matching algorithm - Google Patents
Smart television station caption identification method and system based on gradient direction matching algorithm Download PDFInfo
- Publication number
- CN104598910A CN104598910A CN201510023250.1A CN201510023250A CN104598910A CN 104598910 A CN104598910 A CN 104598910A CN 201510023250 A CN201510023250 A CN 201510023250A CN 104598910 A CN104598910 A CN 104598910A
- Authority
- CN
- China
- Prior art keywords
- station symbol
- gradient direction
- template
- histogram
- matching
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 230000008878 coupling Effects 0.000 claims description 16
- 238000010168 coupling process Methods 0.000 claims description 16
- 238000005859 coupling reaction Methods 0.000 claims description 16
- 238000010586 diagram Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to an intelligent television station caption identification method and system based on a gradient direction matching algorithm, which comprises the following steps: pre-storing a gradient direction template; acquiring the gradient direction of the television station caption; and (3) carrying out gradient direction matching: performing histogram matching; and identifying the television station caption. In order to improve the background interference resistance and remove the influence of irrelevant miscellaneous points, the gradient direction histogram in the traditional statistical region is converted into the gradient direction histogram of the statistical matching points; converting two-stage matching of traditional template matching into multi-stage matching in gradient direction; the method has stronger robustness to background interference, and can be identified only by collecting a single image; and uploading unknown station logos which cannot be identified by the client to the cloud, and updating identification resources by the cloud.
Description
Technical field
The present invention relates to a kind of intelligent television TV station symbol recognition method and system, particularly relate to a kind of intelligent television TV station symbol recognition method and system based on gradient direction matching algorithm.
Background technology
For the TV station symbol recognition under Television clients scene, current mainstream scheme is mainly based on the recognition methods in high in the clouds and the method for classifying modes in conjunction with certain category feature and sorter in this locality.And completely based on the method in high in the clouds, require Television clients accessing Internet to upload real time picture, have larger demand to the network bandwidth and high in the clouds response speed, otherwise recognition result there will be delay; Meanwhile, because station symbol change of background is large, during by the characteristics of image that extracts channel logo position, background characteristics doping wherein, easily causes sorter recognition result mistake.Current mainstream scheme is mainly based on the recognition methods in high in the clouds and the method for classifying modes in conjunction with certain category feature and sorter in this locality.Completely based on the method in high in the clouds, require that Television clients accessing Internet uploads real time picture, have larger demand to the network bandwidth and high in the clouds response speed, otherwise recognition result there will be delay.Because station symbol change of background is large, during by the characteristics of image that extracts channel logo position, background characteristics doping wherein, easily causes sorter recognition result mistake.
Summary of the invention
The technical matters that the present invention solves is: build a kind of intelligent television TV station symbol recognition method and system based on gradient direction matching algorithm, overcomes the technical matters that sorter recognition result mistake easily appears postponing and easily causing in prior art result.
Technical scheme of the present invention is: provide a kind of intelligent television TV station symbol recognition method based on gradient direction matching algorithm, comprise the steps:
Prestore gradient direction template: the template prestoring multiple TV station symbol gradient direction, and the template of described TV station symbol gradient direction comprises the TV station symbol THE TEMPLATE HYSTOGRAM of gradient direction;
Obtain TV station symbol gradient direction: obtain TV station symbol sectional drawing, obtain TV station symbol gradient direction according to TV station symbol sectional drawing;
Carry out gradient direction coupling: division TV station symbol gradient direction being carried out to more than secondary, the TV station symbol gradient direction of GradeNDivision and the template of TV station symbol gradient direction prestored are compared, namely same position point gradient direction is identical thinks this Point matching;
Carry out Histogram Matching: the direction histogram obtaining station symbol match point, calculate TV station symbol match point direction histogram and cling to third constellations distance with the histogram of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores, choose Pasteur apart from minimum be optimum Histogram Matching;
Identify TV station symbol: according to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.
Further technical scheme of the present invention is: carrying out, in gradient direction coupling step, TV station symbol gradient direction being carried out within the scope of 2 π to the division of more than secondary by clockwise sequence.
Further technical scheme of the present invention is: carrying out in Histogram Matching step, and the direction Histogram drawing method obtaining station symbol match point is as follows: the direction adding up all match points, obtains the coupling number of different directions.
Further technical scheme of the present invention is: obtain TV station symbol gradient direction by the following method:
Wherein, dy represents the difference of pixel I (x, y) vertical direction consecutive point chromatic value, and dx represents the difference of pixel I (x, y) horizontal direction consecutive point chromatic value.
Further technical scheme of the present invention is: carry out eight grades of divisions to TV station symbol gradient direction, TV station symbol gradient scope is [0,2 π], to TV station symbol gradient direction by [π/8,3 π/8], [15 π/8,2 π], [13 π/8,15 π/8], [11 π/8,13 π/8], [9 π/8,11 π/8], [7 π/8,9 π/8], [5 π/8,7 π/8], [3 π/8,5 π/8] divide.
Further technical scheme of the present invention is: calculate TV station symbol match point direction histogram as follows with the Histogram distance method of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores:
D
B(p，q)＝-ln(BC(p，q))
Wherein: p represents TV station symbol match point direction histogram, and q represents the TV station symbol THE TEMPLATE HYSTOGRAM that prestores,
Further technical scheme of the present invention is: prestoring in gradient direction template step, comprises and being upgraded the gradient direction template that this locality prestores by network server end.
Further technical scheme of the present invention is: after identification TV station symbol, the TV station symbol gradient direction template of acquisition is uploaded to described server end as new TV station symbol gradient direction template resource.
Technical scheme of the present invention is: build a kind of intelligent television TV station symbol recognition system based on gradient direction matching algorithm, comprise the gradient direction template memory module of the gradient direction template that prestores, obtain the TV station symbol gradient direction acquisition module of TV station symbol gradient direction, carry out the gradient direction matching module of gradient direction coupling, carry out the Histogram Matching module of Histogram Matching, identify TV station's target TV TV station symbol recognition module, described gradient direction template memory module prestores the template of multiple TV station symbol gradient direction, the template of described TV station symbol gradient direction comprises the TV station symbol THE TEMPLATE HYSTOGRAM of gradient direction, described TV station symbol gradient direction acquisition module obtains TV station symbol sectional drawing, obtains TV station symbol gradient direction according to TV station symbol sectional drawing, described gradient direction matching module carries out the division of more than secondary to TV station symbol gradient direction, the TV station symbol gradient direction of GradeNDivision and the template of TV station symbol gradient direction prestored are compared, namely same position point gradient direction is identical thinks this Point matching, described Histogram Matching module obtains the direction histogram of station symbol match point, calculates TV station symbol match point direction histogram and cling to third constellations distance with the histogram of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores, choose Pasteur apart from minimum be optimum Histogram Matching, described TV TV station symbol recognition module is according to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.
Further technical scheme of the present invention is: described gradient direction matching module comprises gradient and divides module, and described gradient divides module carries out more than secondary division to TV station symbol gradient direction, and gradient scope is [0,2 π].
Further technical scheme of the present invention is: described Histogram Matching module comprises Histogram distance computing module, and described Histogram distance computing module calculates Pasteur's distance of TV station symbol match point direction histogram and the TV station symbol THE TEMPLATE HYSTOGRAM that prestores.
Technique effect of the present invention is: the invention provides a kind of intelligent television TV station symbol recognition method and system based on gradient direction matching algorithm: the template prestoring multiple TV station symbol gradient direction, and the template of described TV station symbol gradient direction comprises the TV station symbol THE TEMPLATE HYSTOGRAM of gradient direction; Obtain TV station symbol sectional drawing, obtain TV station symbol gradient direction according to TV station symbol sectional drawing; TV station symbol gradient direction is carried out to the division of more than secondary, the template of the TV station symbol gradient direction of GradeNDivision with the TV station symbol gradient direction prestored is mated; Obtain the histogram in TV station symbol match point direction, calculate TV station symbol match point direction histogram and the Histogram distance of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores, choose optimum Histogram Matching; According to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.The present invention, for improving anti-background interference ability, removes the impact of irrelevant assorted point, traditional statistical regions inside gradient direction histogram is converted to the gradient orientation histogram of statistical match point; The two-stage coupling of conventional template being mated is converted to the multistage coupling of gradient direction; The present invention has stronger robustness for background interference, and only gathering single image can identify; For the unknown station symbol of client None-identified, then be uploaded to high in the clouds, upgrade recognition resource by high in the clouds.
Accompanying drawing explanation
Fig. 1 is that prior art station symbol edge feature is by background interference schematic diagram.
Fig. 2 is process flow diagram of the present invention.
Fig. 3 is TV station symbol gradient direction division figure of the present invention.
Fig. 4 is structural representation of the present invention.
Embodiment
Below in conjunction with specific embodiment, technical solution of the present invention is further illustrated.
As shown in Figure 1, the specific embodiment of the present invention is: provide a kind of intelligent television TV station symbol recognition method based on gradient direction matching algorithm, comprise the steps:
Prestore gradient direction template: the template prestoring multiple TV station symbol gradient direction, and the template of described TV station symbol gradient direction comprises the TV station symbol THE TEMPLATE HYSTOGRAM of gradient direction.
Specific implementation process is as follows:
Obtain TV station symbol gradient direction: obtain TV station symbol sectional drawing, obtain TV station symbol gradient direction according to TV station symbol sectional drawing.Station symbol gradient direction template is stored as sparse matrix, to reduce space size shared by resource bag; The direction histogram of precomputation simultaneously, reduces the real-time calculated amount of client, and stores in the form of vectors.
Obtain TV station symbol gradient direction by the following method:
Wherein, dy represents the difference of pixel I (x, y) vertical direction consecutive point chromatic value, and dx represents the difference of pixel I (x, y) horizontal direction consecutive point chromatic value, and the magnitude range of ori (x, y) is [0,2 π].
As shown in Figure 2, the specific embodiment of the present invention is: division TV station symbol gradient direction being carried out to more than secondary, the TV station symbol gradient direction of GradeNDivision and the template of TV station symbol gradient direction prestored are compared, namely same position point gradient direction is identical thinks this Point matching.In specific implementation process, TV station symbol gradient direction is carried out within the scope of 2 π to the division of more than secondary by clockwise sequence, TV station symbol gradient direction ori (x, y) magnitude range is [0,2 π], be divided into multistage between [0,2 π].In this patent specific embodiment, between [0,2 π], be divided into eight grades, that is: carry out gradient direction 0 ?7 coupling.To TV station symbol gradient direction by [π/8,3 π/8], [15 π/8,2 π], [13 π/8,15 π/8], [11 π/8,13 π/8], [9 π/8,11 π/8], [7 π/8,9 π/8], [5 π/8,7 π/8], [3 π/8,5 π/8] divide.
Carry out Histogram Matching: the direction histogram obtaining station symbol match point, by adding up the direction of all match points, obtaining the coupling number of different directions, namely obtaining the direction histogram of station symbol match point.Calculate the Histogram distance of TV station symbol match point direction histogram and the TV station symbol THE TEMPLATE HYSTOGRAM that prestores, choose Pasteur apart from minimum be optimum Histogram Matching.
Specific implementation process is as follows: calculate TV station symbol match point direction histogram as follows with the Histogram distance method of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores:
D
B(p，q)＝-ln(BC(p，q))
Wherein: p represents TV station symbol match point direction histogram, and q represents the TV station symbol THE TEMPLATE HYSTOGRAM that prestores,
For Bhattacharyya coefficient.
Identify TV station symbol: according to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.The TV station symbol that the TV station symbol gradient direction template that prestores of optimum Histogram Matching is corresponding is current TV station symbol.
The preferred embodiment of the present invention is: prestoring in gradient direction template step, comprise and by network server end, the gradient direction template that this locality prestores being upgraded: loading gradient direction template resource, and compare with the gradient direction template resource of the up-to-date resource version in high in the clouds; If Television clients local resource version falls behind, be then updated to up-to-date resource from high in the clouds; This resource loading procedure loads when local engine initializes.
The preferred embodiment of the present invention is: after identification TV station symbol, the TV station symbol gradient direction template of acquisition is uploaded to described server end as new TV station symbol gradient direction template resource.
Specific implementation process is as follows: if recognition failures and do not exceed uploading pictures restriction, then by picture uploading to high in the clouds pending trial picture library, backstage personnel selection uploads to the new station symbol in pending trial picture library, joins in station symbol template sequence after being marked.High in the clouds starts the new TV station symbol recognition resource of making to the station symbol template base obtained, and issues; Station symbol gradient direction template is stored as sparse matrix, to reduce space size shared by resource bag; The direction histogram of precomputation simultaneously, reduces the real-time calculated amount of client, and stores in the form of vectors.
As shown in Figure 3, the specific embodiment of the present invention is: build a kind of intelligent television TV station symbol recognition system based on gradient direction matching algorithm, comprise the gradient direction template memory module 1 of the gradient direction template that prestores, obtain the TV station symbol gradient direction acquisition module 2 of TV station symbol gradient direction, carry out the gradient direction matching module 3 of gradient direction coupling, carry out the Histogram Matching module 4 of Histogram Matching, identify TV station's target TV TV station symbol recognition module 5, described gradient direction template memory module 1 prestores the template of multiple TV station symbol gradient direction, the template of described TV station symbol gradient direction comprises the TV station symbol THE TEMPLATE HYSTOGRAM of gradient direction, described TV station symbol gradient direction acquisition module 2 obtains TV station symbol sectional drawing and obtains TV station symbol gradient direction according to TV station symbol sectional drawing, described gradient direction matching module 3 pairs of TV station symbol gradient directions carry out the division of more than secondary, the TV station symbol gradient direction of GradeNDivision and the template of TV station symbol gradient direction prestored are compared, namely same position point gradient direction is identical thinks this Point matching, described Histogram Matching module 4 obtains the direction histogram of station symbol match point, calculates TV station symbol match point direction histogram and cling to third constellations distance with the histogram of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores, choose Pasteur apart from minimum be optimum Histogram Matching, described TV TV station symbol recognition module 5 is according to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.
As shown in Figure 3, specific embodiment of the invention process is: gradient direction template memory module 1 obtains TV station symbol gradient direction: obtain TV station symbol sectional drawing, obtains TV station symbol gradient direction according to TV station symbol sectional drawing.
TV station symbol gradient direction acquisition module 2 obtains TV station symbol gradient direction by the following method:
Wherein, dy represents the difference of pixel I (x, y) vertical direction consecutive point chromatic value, and dx represents the difference of pixel I (x, y) horizontal direction consecutive point chromatic value, and the magnitude range of ori (x, y) is [0,2 π].
As shown in Figure 2, described gradient direction matching module 3 comprises gradient and divides module 31, gradient divides the division that module 31 pairs of TV station symbol gradient directions carry out more than secondary, the TV station symbol gradient direction of GradeNDivision and the template of TV station symbol gradient direction prestored compare by described gradient direction matching module 3, and namely same position point gradient direction is identical thinks this Point matching.
Described Histogram Matching module 4 obtains the direction histogram of station symbol match point, by adding up the direction of all match points, obtaining the coupling number of different directions, namely obtaining the direction histogram of station symbol match point.Described Histogram Matching module 4 comprises Histogram distance computing module 41, and described Histogram distance computing module 41 calculates TV station symbol match point direction histogram and Pasteur's distance of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores.Histogram distance computing module 41 calculates TV station symbol match point direction histogram and prestores the Histogram distance of TV station symbol THE TEMPLATE HYSTOGRAM, choose Pasteur apart from minimum be optimum Histogram Matching.
Specific implementation process is as follows: calculate TV station symbol match point direction histogram as follows with the Histogram distance method of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores:
D
B(p，q)＝-ln(BC(p，q))
Wherein: p represents TV station symbol match point direction histogram, and q represents the TV station symbol THE TEMPLATE HYSTOGRAM that prestores,
For Bhattacharyya coefficient.
TV TV station symbol recognition module 5 is according to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.The TV station symbol that the TV station symbol gradient direction template that prestores of optimum Histogram Matching is corresponding is current TV station symbol.
In specific implementation process, described gradient divides module 31 pairs of TV station symbol gradient directions carry out more than secondary within the scope of 2 π division by clockwise sequence, TV station symbol gradient direction ori (x, y) magnitude range is [0,2 π], be divided into multistage between [0,2 π].In this patent specific embodiment, [0,2 π] between be divided into eight grades: to TV station symbol gradient direction by [π/8,3 π/8], [15 π/8,2 π], [13 π/8,15 π/8], [11 π/8,13 π/8], [9 π/8,11 π/8], [7 π/8,9 π/8], [5 π/8,7 π/8], [3 π/8,5 π/8] divide.
Technique effect of the present invention is: the invention provides a kind of intelligent television TV station symbol recognition method and system based on gradient direction matching algorithm: the template prestoring multiple TV station symbol gradient direction, and the template of described TV station symbol gradient direction comprises the TV station symbol THE TEMPLATE HYSTOGRAM of gradient direction; Obtain TV station symbol sectional drawing, obtain TV station symbol gradient direction according to TV station symbol sectional drawing; TV station symbol gradient direction is carried out to the division of more than secondary, the template of the TV station symbol gradient direction of GradeNDivision with the TV station symbol gradient direction prestored is mated; Obtain the histogram in TV station symbol match point direction, calculate TV station symbol match point direction histogram and the Histogram distance of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores, choose optimum Histogram Matching; According to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.The present invention, for improving anti-background interference ability, removes the impact of irrelevant assorted point, traditional statistical regions inside gradient direction histogram is converted to the gradient orientation histogram of statistical match point; The two-stage coupling of conventional template being mated is converted to the multistage coupling of gradient direction; The present invention has stronger robustness for background interference, and only gathering single image can identify; For the unknown station symbol of client None-identified, then be uploaded to high in the clouds, upgrade recognition resource by high in the clouds.
Above content is in conjunction with concrete preferred implementation further description made for the present invention, can not assert that specific embodiment of the invention is confined to these explanations.For general technical staff of the technical field of the invention, without departing from the inventive concept of the premise, some simple deduction or replace can also be made, all should be considered as belonging to protection scope of the present invention.
Claims (11)
1., based on an intelligent television TV station symbol recognition method for gradient direction matching algorithm, comprise the steps:
Prestore gradient direction template: the template prestoring multiple TV station symbol gradient direction, and the template of described TV station symbol gradient direction comprises the TV station symbol THE TEMPLATE HYSTOGRAM of gradient direction;
Obtain TV station symbol gradient direction: obtain TV station symbol sectional drawing, obtain TV station symbol gradient direction according to TV station symbol sectional drawing;
Carry out gradient direction coupling: division TV station symbol gradient direction being carried out to more than secondary, the TV station symbol gradient direction of GradeNDivision and the template of TV station symbol gradient direction prestored are compared, namely same position point gradient direction is identical thinks this Point matching;
Carry out Histogram Matching: the direction histogram obtaining station symbol match point, calculate TV station symbol match point direction histogram and cling to third constellations distance with the histogram of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores, choose Pasteur apart from minimum be optimum Histogram Matching;
Identify TV station symbol: according to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.
2. according to claim 1 based on the intelligent television TV station symbol recognition method of gradient direction matching algorithm, it is characterized in that, carrying out, in gradient direction coupling step, TV station symbol gradient direction being carried out within the scope of 2 π to the division of more than secondary by clockwise sequence.
3. according to claim 1 based on the intelligent television TV station symbol recognition method of gradient direction matching algorithm, it is characterized in that, carrying out in Histogram Matching step, the direction Histogram drawing method obtaining station symbol match point is as follows: the direction adding up all match points, obtains the coupling number of different directions.
4. according to claim 1 based on the intelligent television TV station symbol recognition method of gradient direction matching algorithm, it is characterized in that, obtain TV station symbol gradient direction by the following method:
Wherein, dy represents the difference of pixel I (x, y) vertical direction consecutive point chromatic value, and dx represents the difference of pixel I (x, y) horizontal direction consecutive point chromatic value.
5. according to claim 1 based on the intelligent television TV station symbol recognition method of gradient direction matching algorithm, it is characterized in that, eight grades of divisions are carried out to TV station symbol gradient direction, TV station symbol gradient scope is [0,2 π], to TV station symbol gradient direction by [π/8,3 π/8], [15 π/8,2 π], [13 π/8,15 π/8], [11 π/8,13 π/8], [9 π/8,11 π/8], [7 π/8,9 π/8], [5 π/8,7 π/8], [3 π/8,5 π/8] divide.
6. according to claim 1 based on the intelligent television TV station symbol recognition method of gradient direction matching algorithm, it is characterized in that, calculate TV station symbol match point direction histogram as follows with the Histogram distance method of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores:
D
B(p，q)＝-ln(BC(p，q))
Wherein: p represents TV station symbol match point direction histogram, and q represents the TV station symbol THE TEMPLATE HYSTOGRAM that prestores,
7., according to claim 1 based on the intelligent television TV station symbol recognition method of gradient direction matching algorithm, it is characterized in that, prestoring in gradient direction template step, comprise and by network server end, the gradient direction template that this locality prestores being upgraded.
8. according to claim 1 based on the intelligent television TV station symbol recognition method of gradient direction matching algorithm, it is characterized in that, after identification TV station symbol, the TV station symbol gradient direction template of acquisition is uploaded to described server end as new TV station symbol gradient direction template resource.
9. the intelligent television TV station symbol recognition system based on gradient direction matching algorithm, it is characterized in that, comprise the gradient direction template memory module of the gradient direction template that prestores, obtain the TV station symbol gradient direction acquisition module of TV station symbol gradient direction, carry out the gradient direction matching module of gradient direction coupling, carry out the Histogram Matching module of Histogram Matching, identify TV station's target TV TV station symbol recognition module, described gradient direction template memory module prestores the template of multiple TV station symbol gradient direction, the template of described TV station symbol gradient direction comprises the TV station symbol THE TEMPLATE HYSTOGRAM of gradient direction, described TV station symbol gradient direction acquisition module obtains TV station symbol sectional drawing, obtains TV station symbol gradient direction according to TV station symbol sectional drawing, described gradient direction matching module carries out the division of more than secondary to TV station symbol gradient direction, the TV station symbol gradient direction of GradeNDivision and the template of TV station symbol gradient direction prestored are compared, namely same position point gradient direction is identical thinks this Point matching, described Histogram Matching module obtains the direction histogram of station symbol match point, calculates TV station symbol match point direction histogram and cling to third constellations distance with the histogram of the TV station symbol THE TEMPLATE HYSTOGRAM that prestores, choose Pasteur apart from minimum be optimum Histogram Matching, described TV TV station symbol recognition module is according to optimum Histogram Matching and the current TV station symbol of the TV station symbol gradient direction template identification that prestores.
10. according to claim 9 based on the intelligent television TV station symbol recognition system of gradient direction matching algorithm, it is characterized in that, described gradient direction matching module comprises gradient and divides module, described gradient divides module carries out more than secondary division to TV station symbol gradient direction, gradient scope is [0,2 π].
11. according to claim 9 based on the intelligent television TV station symbol recognition system of gradient direction matching algorithm, it is characterized in that, described Histogram Matching module comprises Histogram distance computing module, and described Histogram distance computing module calculates Pasteur's distance of TV station symbol match point direction histogram and the TV station symbol THE TEMPLATE HYSTOGRAM that prestores.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510023250.1A CN104598910A (en) | 2015-01-16 | 2015-01-16 | Smart television station caption identification method and system based on gradient direction matching algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510023250.1A CN104598910A (en) | 2015-01-16 | 2015-01-16 | Smart television station caption identification method and system based on gradient direction matching algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104598910A true CN104598910A (en) | 2015-05-06 |
Family
ID=53124682
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510023250.1A Pending CN104598910A (en) | 2015-01-16 | 2015-01-16 | Smart television station caption identification method and system based on gradient direction matching algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104598910A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017113605A1 (en) * | 2015-12-28 | 2017-07-06 | 深圳Tcl数字技术有限公司 | Method and device for channel identification |
| CN111079802A (en) * | 2019-12-02 | 2020-04-28 | 易思维（杭州）科技有限公司 | Matching method based on gradient information |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100007797A1 (en) * | 2008-07-08 | 2010-01-14 | Zeitera, Llc | Digital Video Fingerprinting Based on Resultant Weighted Gradient Orientation Computation |
| CN101739561A (en) * | 2008-11-11 | 2010-06-16 | 中国科学院计算技术研究所 | TV station logo training method and identification method |
| CN103473551A (en) * | 2013-09-16 | 2013-12-25 | 中国传媒大学 | Station logo recognition method and system based on SIFT operators |
-
2015
- 2015-01-16 CN CN201510023250.1A patent/CN104598910A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100007797A1 (en) * | 2008-07-08 | 2010-01-14 | Zeitera, Llc | Digital Video Fingerprinting Based on Resultant Weighted Gradient Orientation Computation |
| CN101739561A (en) * | 2008-11-11 | 2010-06-16 | 中国科学院计算技术研究所 | TV station logo training method and identification method |
| CN103473551A (en) * | 2013-09-16 | 2013-12-25 | 中国传媒大学 | Station logo recognition method and system based on SIFT operators |
Non-Patent Citations (2)
| Title |
|---|
| 章金华: "电视台台标识别与检索方法研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 胡星火: "基于改进Chamfer匹配的台标识别", 《计算机工程》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017113605A1 (en) * | 2015-12-28 | 2017-07-06 | 深圳Tcl数字技术有限公司 | Method and device for channel identification |
| CN105635750B (en) * | 2015-12-28 | 2018-07-03 | 深圳Tcl数字技术有限公司 | Channel recognition methods and device |
| CN111079802A (en) * | 2019-12-02 | 2020-04-28 | 易思维（杭州）科技有限公司 | Matching method based on gradient information |
| CN111079802B (en) * | 2019-12-02 | 2023-04-07 | 易思维（杭州）科技有限公司 | Matching method based on gradient information |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Huang et al. | Building extraction from multi-source remote sensing images via deep deconvolution neural networks | |
| CN109416413B (en) | Solar energy forecast | |
| CN110036392A (en) | System and method for being drawn based on multiple-pass data | |
| CN108089191B (en) | Global positioning system and method based on laser radar | |
| CN108369651B (en) | Method, system and non-transitory computer-readable storage medium for extracting sky area | |
| CN109784205B (en) | Intelligent weed identification method based on multispectral inspection image | |
| CN102881160B (en) | Outdoor traffic sign identification method under low-illumination scene | |
| CN104463200A (en) | Satellite remote sensing image sorting method based on rule mining | |
| CN103984963B (en) | Method for classifying high-resolution remote sensing image scenes | |
| CN105335469A (en) | Method and device for image matching and retrieving | |
| CN106951863B (en) | Method for detecting change of infrared image of substation equipment based on random forest | |
| CN113252584B (en) | Crop growth detection method and system based on 5G transmission | |
| CN105741309A (en) | Remote sensing image change detection method based on chi-square transformation and sample selection | |
| CN112862150A (en) | Forest fire early warning method based on image and video multi-model | |
| CN104978418A (en) | Intelligent weather image acquisition and identification method and system | |
| CN104598910A (en) | Smart television station caption identification method and system based on gradient direction matching algorithm | |
| CN108073865B (en) | Aircraft trail cloud identification method based on satellite data | |
| CN104036294A (en) | Spectral tag based adaptive multi-spectral remote sensing image classification method | |
| CN115409789A (en) | Power transmission line engineering defect detection method based on image semantic segmentation | |
| CN109919936B (en) | Method, device and equipment for analyzing running state of composite insulator | |
| CN104185069B (en) | A kind of TV station symbol recognition method and its identifying system | |
| CN101876993A (en) | Method for extracting and retrieving textural features from ground digital nephograms | |
| CN104517127A (en) | Self-learning pedestrian counting method and apparatus based on Bag-of-features model | |
| CN105354591A (en) | High-order category-related prior knowledge based three-dimensional outdoor scene semantic segmentation system | |
| CN105205485A (en) | Large scale image segmentation algorithm based on multi-inter-class maximum variance algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150506 |
|
| RJ01 | Rejection of invention patent application after publication |