CN1567978A - Map border category sensing method with noise filtering step - Google Patents
Map border category sensing method with noise filtering step Download PDFInfo
- Publication number
- CN1567978A CN1567978A CN 03148556 CN03148556A CN1567978A CN 1567978 A CN1567978 A CN 1567978A CN 03148556 CN03148556 CN 03148556 CN 03148556 A CN03148556 A CN 03148556A CN 1567978 A CN1567978 A CN 1567978A
- Authority
- CN
- China
- Prior art keywords
- pectination
- factor
- picture frame
- counter
- values
- 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
Links
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention is a frame class detecting method with noise filtering step, it carries on a denoising step after calculating the honeycomb factor, in order to confirm if it belongs to honeycomb factor or noise point, finally carries on the judgment of frame class, reduces the probability judging the noise point as comb factor, and achieves the aim for upgrading the accuracy of frame class judgment.
Description
Technical field
The present invention is a kind of picture frame kind method for detecting of tool noise filtering step.Be after calculating the pectination factor, carry out a noise filtering (Denoising) step, improve the purpose of judging picture frame kind accuracy to reach.
Background technology
The image of video pictures is made up of per second 30 picture frames of transmission (frame) at present, and then produces the continuous dynamic image, and each picture frame is to be combined (with National Television System Committee by the multi-strip scanning line; NTSC is an example, is to transmit 525 scan lines), just each picture frame is to utilize 1/30 second to transmit 525 scan lines.
And the picture frame of above-mentioned video signal can be divided into two kinds at present, can be divided into alternating expression picture frame (interlace) and (progressive) picture frame in turn respectively, wherein its scan mode of alternating expression picture frame is that each picture frame is further divided into figure field, top (top field) and base map field (bottom field), wherein top figure is the scan line of odd number, the base map field then is the scan line of even number, be divided into twice scanning respectively then, after scanning the base map field again behind the elder generation scanning figure field, top, produce a complete picture frame; The scan mode of (progressive) picture frame is then for once transmitting simultaneously all scan lines of whole picture frame (frame), a line one line sequential scanning then in turn.
Shown in Fig. 1 (a) and Fig. 1 (b), be respectively alternating expression picture frame 10 and reach the image schematic diagram of picture frame 11 in turn, by as can be known aforementioned, because the scan mode of alternating expression picture frame 10 is for to transmit figure field, top (odd-numbered scan lines) and base map field (even-line interlace line) respectively in the different time, therefore can as the square image 100 among Fig. 1 (a) and 101 its contour edges generations of triangle image as the image (being commonly referred to as the pectination factor (comb factor)) of pectination, resolution is relatively poor, and relative picture quality is also relatively poor; And because picture frame 11 is because for transmitting whole picture frame simultaneously in turn, so it can not produce as the pectination factor among Fig. 1 (a).
By aforementioned as can be known because the generation of the pectination factor has great influence for resolution, therefore before image output, can carry out a detecting flow process, be staggered picture frame or picture frame in turn with the picture frame of detecting input, if detecting is staggered picture frame, then work as picture and before output, have a release of an interleave (deinterlacing) treatment step so that the pectination factor is removed, to obtain better image output quality; If picture frame in turn, then need be through the release of an interleave step and directly image output get final product.
Traditional detecting flow process as shown in Figure 2, at first in step 201, if counter is zero, input one can calculate the picture frame letter formula F (n) of picture element pectination factor values (comb factor), then carry out step 202, in this picture frame letter formula F (n) of all picture element substitutions, step 203, calculate picture element pectination factor values (comb factor), the calculation formula that this pectination factor values is calculated is narrated in the back; In the step 204, whether judge the pectination factor values,, then enter in the step 205, count value is added 1 if greater than critical value greater than critical value; If the pectination factor values less than critical value, then enters step 206, judge whether to be last pectination factor; If then enter step 207; If not, then repeating step 203 to step 206; Step 207, whether judge in the counter numerical value less than critical value,, then judge that in step 208 its picture frame kind is picture frame in turn if when numerical value is less than critical value in the counter, if during greater than critical value, then entering step 209, counter values judges that its picture frame kind is the alternating expression picture frame; At last, finish the detecting flow process.
It calculates the picture frame picture element schematic diagram of its pectination factor values of a picture element to Figure 3 shows that in Fig. 2 step 203, wherein desire is calculated the position that picture element 30 betides the X shown in the figure, and shown in zero the first contiguous picture element 31 be picture element arround contiguous with the second contiguous picture element 32, calculation formula of its calculating picture element 30 pectination factor values is:
Pectination factor values (Comb Factor (x, y))=(b-x) * (e-x)-(b-e)
2
Wherein, x is that picture element 30 is in coordinate (x, picture element value y); B is that the first contiguous picture element 31 is in coordinate (x, picture element value y-1); E is that the second contiguous picture element 32 is in coordinate (x, picture element value y+1).
Though utilizing numerical value that the calculation formula calculates each pectination factor to make comparisons with a fixedly critical value of setting then can be used as one and judges whether this picture element is the pectination factor, at last can judge the picture frame kind according to the number (being the counter number) of the pectination factor again, but this kind method also can cause erroneous judgement.Its reason be since when a picture element as if with arround the luminance difference of picture element when very big, its pectination factor values of utilizing above-mentioned calculation formula to be calculated can be greatly, therefore, if when on an image, having at assorted, its also might because of with arround the luminance difference of picture element greatly be mistaken for the pectination factor, for example if when one has at more assorted in the image in turn, promptly can to cause last erroneous judgement be the alternating expression picture frame because of being judged as the pectination factor.
Above-mentioned for solving because of the erroneous judgement of the assorted point on the image be picture frame kind method for detecting that the problem of the pectination factor, the present invention propose a kind of tool noise filtering step reaches more accurate alternating expression picture frame and reaches picture frame determination methods in turn.
Summary of the invention
The present invention is a kind of picture frame kind method for detecting of tool noise filtering step, be after calculating the pectination factor, carry out a noise filtering (Denoising) step, to confirm that it belongs to the pectination factor or assorted point, the last step of judging the picture frame kind again, so can calculate the number of the pectination factor accurately, reducing will mix by mistake a little is judged as the possibility of the pectination factor, improves the purpose of judging picture frame kind accuracy to reach.
For achieving the above object, the picture frame kind method for detecting of tool noise filtering step of the present invention is input picture frame one letter formula F (n) in processor earlier, and to establish counter be zero; Bring into all picture elements in the picture frame in this letter formula then and judge whether each picture element is the pectination factor; Then, whether serve as true or be the point of mixing to judge each pectination factor through a noise filtering step again; If the pectination factor then adds one with counter; At last, judge that according to the counter number it is alternating expression picture frame or picture frame in turn.
Description of drawings
Relevant detailed content of the present invention and technology, existing conjunction with figs. is described as follows:
Fig. 1 (a) is the image output schematic diagram of alternating expression picture frame;
Fig. 1 (b) is the image output schematic diagram of picture frame in turn;
Fig. 2 is the schematic flow sheet of prior art detecting picture frame kind class methods;
Fig. 3 is for calculating the picture frame picture element schematic diagram of its pectination factor values of a picture element;
Fig. 4 is one of its preferred embodiment of picture frame kind method for detecting of a tool noise filtering step of the present invention schematic flow sheet; And
Fig. 5 is the schematic flow sheet of its another preferred embodiment of picture frame kind method for detecting of tool noise filtering step of the present invention.
Symbol description:
10 alternating expression picture frames; 100 square images; 101 triangle images;
11 picture frames in turn; 30 picture elements; 31 first contiguous picture elements;
32 second contiguous picture elements;
301,401 inputs, one letter formula F (n) is to processing, and establishing counter is zero;
302,402 all picture element substitutions this letter formula F (n) are calculated the pectination factor values;
303,403 pectinations factor values＞pectination factor critical value;
304,404 first pectination figure are 1;
305,405 first pectination figure are zero;
306,406 last pectination factor values;
307,407 carry out the noise filtering step;
308,408 calculate the judgment value of each picture element;
309,409 judgment value＞most critical values;
310,413 calculators add 1;
311 last judgment value; 414 last second pectination figure;
312,415 counter values＜counter critical value;
313,416 picture frames in turn;
314,417 alternating expression picture frames.
Embodiment
Fig. 4 is a flow process schematic diagram of the picture frame kind method for detecting of a preferred embodiment of the present invention tool noise filtering step, comprises the following steps:
At first, carry out step 301, establishing counter is zero, imports a picture frame letter formula F (n) to a processor, owing to this picture frame letter formula is mentioned in the prior art, this no longer superfluous chatting; Then, in step 302,, and calculate the pectination factor values with among all picture element substitutions this letter formula F (n); Then, whether in step 303, judge the pectination factor values,, then enter step 304 value of the first pectination figure (Comb map1) of its correspondence is made as 1 if during greater than pectination factor critical value greater than pectination factor critical value; If less than pectination factor critical value, then in step 305, its value of the first pectination figure (Comb map1) of correspondence is made as zero; Then carry out step 306, judge whether pectination factor graph, if the pectination factor graph of last picture element into last picture element, then enter step 307, if not the pectination factor graph of last picture element then enters repeating step 302 to 305, till the pectination factor graph that calculates last picture element.
Then carry out step 307, carry out noise filtering step (Median Flitering), it is the pairing first pectination figure of each picture element to be brought into to calculate a judgment value (counts) in the following calculation formula in step 308 again, and its calculation formula is as follows at present embodiment:
Judgment value
Wherein, (j=-1～1 is all one 3 * 3 matrix for x+i, y+j) for i=-1～1, and (x+i then is that picture element out of the ordinary is in correspondence position (x+i, the first pectination figure value that y+j) had before been calculated y+j) to Comb map for M and Combmap.
After calculating this judgment value, then in step 309, judgment value and the most critical values (Majority-th) calculated are done a relatively judgement, if greater than most critical values, then enter step 310 counter is added one; If this judgment value is less than most critical values (Majority-th), then enters step 311 and judge whether judgment value, if then enter step 312 into last picture element; If not then repeating step 308 to 309.
In step 312, judge that whether counter values is less than a counter critical value, the numerical value of its counter is promptly represented the number of the pectination factor, if counter values is less than the counter critical value, represent pectination factor number in this picture frame not reach the standard of alternating expression picture frame, judge that in step 312 it is picture frame in turn; If greater than the counter critical value, represent the pectination factor number in this picture frame to surpass the standard value that its pectination factor number of alternating expression picture frame is set, enter step 313 and judge that it is the alternating expression picture frame, last, finish the detecting flow process.
Be illustrated in figure 5 as the schematic flow sheet of another preferred embodiment of the present invention, itself and last embodiment main difference are to increase by one the make comparisons result that judges of judgment value and most critical value (Majority-th) are inputed to the step of one second pectination figure, and it comprises the following steps:
At first, carry out step 401, establishing counter is zero, imports a picture frame letter formula F (n) to a processor, owing to this picture frame letter formula is mentioned in the prior art, this no longer superfluous chatting; Then, in step 402,, and calculate the pectination factor values with among all picture element substitutions this letter formula F (n); Then, whether in step 403, judge the pectination factor values,, then enter step 404 its value of the first pectination figure (Comb map1) of its correspondence is made as 1 if during greater than pectination factor critical value greater than pectination factor critical value; If less than pectination factor critical value, then in step 405, its value of the first pectination figure (Comb map1) of correspondence is made as zero; Then carry out step 406, judge whether pectination factor graph, if the pectination factor graph of last picture element into last picture element, then enter step 407, if not the pectination factor graph of last picture element then enters repeating step 402 to 405, till the pectination factor graph that calculates last picture element.
Then carry out step 407, carry out noise filtering step (Median Flitering), at first in step 408 the pairing first pectination figure of each picture element brought into and calculate a judgment value (counts) in the following calculation formula, its calculation formula is as follows at present embodiment:
Judgment value
Wherein, (x+i, y+j) for i=-1～1, j=-1～1 are all one 3 * 3 Matrix C ombmap (x+i then are that picture element out of the ordinary is in correspondence position (x+i, the first pectination figure value of y+j) before having calculated y+j) for M and Combmap.
After calculating this judgment value, then in step 409, judgment value and the most critical values (Majority-th) calculated are done a relatively judgement, and discrimination result is imported among the one second pectination figure, be a pectination figure (bit map) at this pectination figure.This step is to utilize judgment value that aforementioned calculation formula calculates in order to judge that this picture element is the still assorted point of the pectination factor, and in the present embodiment, to establish its second pectination figure value be 1 if the judgment value of this picture element, then enters step 410 greater than most critical values; If this judgment value is less than most critical values, then entering step 411 is zero with the value of the second pectination figure of correspondence.
In step 412, judge whether each second pectination figure is 1; If then counter is added 1; If not, then enter and judge whether to be last second pectination figure in the step 414, if, then enter in the step 415, if not, then return step 412.Step 415 is for to judge that whether counter values is less than the counter critical value, the numerical value of its counter is promptly represented the number of the pectination factor, if counter values is less than the counter critical value, represent pectination factor number in this picture frame not reach the standard of alternating expression picture frame, judge that then it is picture frame (step 416) in turn; If on behalf of the pectination factor number in this picture frame, counter values greater than the counter critical value, surpassed the standard value that its pectination factor number of alternating expression picture frame is set, judge that then it is an alternating expression picture frame (step 417), last, finish the detecting flow process.
It more than is the detailed description of the picture frame kind method for detecting embodiment of tool noise filtering step of the present invention, after calculating the pectination factor, utilize a noise filtering (Denoising) step, to confirm that its picture element corresponding to this pectination factor belongs to the pectination factor (Comb factor) or assorted point, the last step of judging the picture frame kind again, so can calculate the number of the pectination factor in the picture frame accurately, reducing will mix by mistake a little is judged as the possibility of the pectination factor, improves the purpose of judging picture frame kind accuracy to reach.
In sum, fully demonstrate picture frame kind method for detecting all dark well-off progressive of executing on purpose and effect of tool noise filtering step of the present invention, has the value on the industry, and be present new invention not seen before on the market, meet the condition of patent of invention fully, so file an application in accordance with the law.
The front is narrated, and only is preferred embodiment of the present invention, not as limiting the scope that the present invention implemented.Promptly the equalization of being made according to the present patent application claim generally changes and modifies, and all should still belong in the scope that patent of the present invention contains.
Claims (7)
1, a kind of picture frame kind method for detecting of tool noise filtering step, it is finished by a processor, stores the letter formula that can calculate the pectination factor values in this processor, it is characterized in that this method for detecting includes:
Picture element in one picture frame is inputed in the processor in regular turn;
Calculate the pectination factor values of its all picture elements of picture frame according to this letter formula;
Carry out a noise filtering step, and recognize this respectively pectination factor values of this picture element;
If belong to the pectination factor, then counter is added one;
If do not belong to the pectination factor, then counter is constant; And
Judge this picture frame kind according to counter values and counter critical value; And
Finish the detecting flow process.
2, the picture frame kind method for detecting of tool noise filtering step as claimed in claim 1 is characterized in that comprising in carrying out a noise filtering step:
A this pectination factor and a pectination factor critical value are done one relatively;
If during greater than pectination factor critical value, then the first pectination figure with its correspondence is made as one;
If less than pectination factor critical value, then the first pectination figure with correspondence is made as zero; And
Calculate the respectively judgment value of its first pectination of this picture element figure.
3, the picture frame kind method for detecting of tool noise filtering step as claimed in claim 2 is characterized in that also comprising:
Judgment value and most critical values are done one relatively,, then counter is added one if greater than most critical values; And
If less than most critical values, counter remains unchanged.
4, the picture frame kind method for detecting of tool noise filtering step as claimed in claim 2 is characterized in that also comprising:
Judgment value and most critical values are done one relatively, if greater than most critical values, then establishing the second corresponding pectination figure is one; If it is zero that this judgment value, is then established the second corresponding pectination figure less than most critical values; And
Judge whether each second pectination figure is one; If then counter is added one; If not then counter remains unchanged.
5, the picture frame kind method for detecting of tool noise filtering step as claimed in claim 1, it is characterized in that: judging according to counter values in the step of this picture frame kind, if counter values is greater than this counter critical value, then determine that it is picture frame in turn, if less than this counter critical value, then determine that it is the alternating expression picture frame.
6, a kind of picture frame kind method for detecting of tool noise filtering step, it is finished by a processor, stores the letter formula that can calculate the pectination factor values in this processor, it is characterized in that this method for detecting includes:
Picture element in one picture frame is inputed in the processor in regular turn;
Calculate the pectination factor values of all picture elements according to this picture frame letter formula;
Calculate the respectively first pectination figure value of this picture element;
According to the first pectination figure value of this picture element respectively, calculate the respectively judgment value of this picture element, and judgment value and most critical values are done one relatively, if greater than most critical values, then establishing the second corresponding pectination figure is one; If it is zero that this judgment value, is then established the second corresponding pectination figure less than most critical values; And
Judge whether each second pectination figure is one; If then counter is added one; If not then counter remains unchanged.
Judge this picture frame kind according to counter values and counter critical value; And
Finish the detecting flow process.
7, the picture frame kind method for detecting of tool noise filtering step as claimed in claim 6, it is characterized in that: in calculating the step of the first pectination figure value of this picture element respectively, a this pectination factor values and a pectination factor critical value are done one relatively, if during greater than pectination factor critical value, then the first pectination figure value with its correspondence is made as one; If less than pectination factor critical value, then the first pectination figure value with correspondence is made as zero.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031485561A CN100334874C (en) | 2003-07-02 | 2003-07-02 | Map border category sensing method with noise filtering step |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031485561A CN100334874C (en) | 2003-07-02 | 2003-07-02 | Map border category sensing method with noise filtering step |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1567978A true CN1567978A (en) | 2005-01-19 |
| CN100334874C CN100334874C (en) | 2007-08-29 |
Family
ID=34472312
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB031485561A Expired - Fee Related CN100334874C (en) | 2003-07-02 | 2003-07-02 | Map border category sensing method with noise filtering step |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100334874C (en) |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FI980132A (en) * | 1998-01-21 | 1999-07-22 | Nokia Mobile Phones Ltd | Adaptive post-filter |
| US7110455B2 (en) * | 2001-08-14 | 2006-09-19 | General Instrument Corporation | Noise reduction pre-processor for digital video using previously generated motion vectors and adaptive spatial filtering |
-
2003
- 2003-07-02 CN CNB031485561A patent/CN100334874C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN100334874C (en) | 2007-08-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1266934C (en) | Method and device for video frequency removing interleave | |
| CN1290057C (en) | Improved spatial resolution of video images | |
| CN1130069C (en) | False contour correcting apparatus and method | |
| CN101042735A (en) | Image binarization method and device | |
| CN1812553A (en) | Method of edge based pixel location and interpolation | |
| CN1317900C (en) | Apparatus and method for detecting a 2:2 pull-down sequence | |
| CN1633159A (en) | A method for removing image noise | |
| CN1941923A (en) | Automatic white balance method for color digital image | |
| CN101029823A (en) | Method for tracking vehicle based on state and classification | |
| CN1155225C (en) | Mobile detecting circuit and noise suppression circuit containing the same circuit | |
| CN1758744A (en) | Can select the image processing apparatus and the method thereof of field | |
| CN106846815A (en) | A kind of efficient intelligent traffic administration system big data analysis system | |
| CN1266943C (en) | Device and process for estimating noise level, noise reduction system and coding system comprising such a device | |
| CN1293756C (en) | Video signal processing apparatus, video signal processing method and video source determining method | |
| CN1525387A (en) | Device and method for detecting blurring of image | |
| CN1567978A (en) | Map border category sensing method with noise filtering step | |
| CN1941912A (en) | Image transmission system and method for analyzing fluctuation zone during image transmission | |
| CN1595957A (en) | Method for determining automatic detection threshold of bad pixel of medical image | |
| CN1738431A (en) | Frame field self-adaptive detection method | |
| CN1741592A (en) | Can detect wrong editor's film mode detection apparatus and method thereof | |
| CN1825941A (en) | Pull-down detection apparatus and pull-down detection method | |
| CN1469303A (en) | Apparatus for conducting binary coding to image and using method thereof | |
| CN1301017C (en) | Method for testing variety of frames with adjustable critical values | |
| CN1212732C (en) | Digital image edge interpolation method | |
| CN100337463C (en) | Method for searching black and white boundary of image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20070829 Termination date: 20160702 |