Help for image pattern matching

Hello Everyone
I am working for my last year project. In my project I will work on the image processing to find a moving object. I will work by JMF. I have finished to grab a frame from the webcam video clips. Now I need a algorithm to find a Image pattern from the grabed image. But I donot know which algorithm is fine for image pattern matching as well as how can I implement in java. Is anyone know please help me very urgently.
Thank you
Md. Mainul Hasan

If you would like to take a look at http://www.exactfutures.com/index01.htm and http://www.exactfutures.com/index02.htm and http://www.codeproject.com/useritems/activity.asp then these pages and links may well be useful to you. It may not be exactly what you are looking for, but it does point to some examples with source for video analytics, and at the very least they illustrate how to capture & handle the data including a fast movement detection algorithm. If you want to find a specific shape then search the internet for information on chamfer distance transforms - one can use JMF or extend these simple examples to apply those techniques.

Similar Messages

Retain Pattern Match Attribute for Next Pattern Match Calculation

When a MATCH_RECOGNIZE PATTERN match occurs, is there any way to retain the underlying event for the next pattern match?
It's like I want to replay the underlying event (that triggered the match) into the next "batch" of events.

Here is an approach
1) Use a (local) cache to hold the "lastMatchedTime"
2) Augment each input event with the "lastMatchedTime" by first (left outer join) joining with cache
3) Run the MATCH_RECOGNIZE on the augmented stream
4) Use a match to update the cache
Here are the queries that I tried -
<view id="S">
        <![CDATA[
             RSTREAM (
                SELECT
                       id,
                       obs,
                       epoch,
                       nvl(C.lastMatchedTime, 0) as lastMatchedTime
                FROM
                       InputChannel[now] AS I LEFT OUTER JOIN lastMatchCache AS C
                ON
                       C.keyValue = "key"
        ]]>
      </view>
      <view id="v1">
        <![CDATA[
                  SELECT
                          T.id,
                          T.obsTotal,
                          T.elapsedTime,
                          T.firstts,
                          T.lastts
                  FROM S MATCH_RECOGNIZE (
                          PARTITION BY id
                          MEASURES
                                    id as id,
                                    sum(obs) as obsTotal,
                                    B.epoch - decode(B.lastMatchedTime, 0L, min(epoch), B.lastMatchedTime) as elapsedTime,
                                    decode(B.lastMatchedTime, 0L, min(epoch), B.lastMatchedTime) as firstts,
                                    B.epoch as lastts
                          PATTERN ( A+? B )
                          DEFINE
                                   B AS (epoch- decode(lastMatchedTime, 0L, min(epoch), lastMatchedTime)) > 20*60*1000
                  ) as T
        ]]>
      </view>
      <query id="qcache">
         <![CDATA[
             SELECT
                   "key" as keyValue,
                   lastts as lastMatchedTime
             FROM
                   v1
         ]]>
      </query>
       <query id="q1">
         <![CDATA[
             SELECT
             FROM
                    v1
         ]]>
      </query>
    Also, include in the config file
<channel>
       <name>CacheChannel</name>
       <selector>qcache</selector>
   </channel>
   <channel>
       <name>OutputChannel</name>
       <selector>q1</selector>
   </channel>The EPN that I used was
<wlevs:event-type-repository>
        <wlevs:event-type type-name="CacheEvent">
            <wlevs:class>com.bea.wlevs.event.example.helloworld.CacheEvent</wlevs:class>
        </wlevs:event-type>
        <wlevs:event-type type-name="InputEvent">
           <wlevs:properties>
              <wlevs:property name="id" type="int"/>
              <wlevs:property name="obs" type="int"/>
              <wlevs:property name="epoch" type="bigint"/>
           </wlevs:properties>
        </wlevs:event-type>
        <wlevs:event-type type-name="OutputEvent">
           <wlevs:properties>
              <wlevs:property name="id" type="int"/>
              <wlevs:property name="obsTotal" type="int"/>
              <wlevs:property name="elapsedTime" type="bigint"/>
              <wlevs:property name="firstts" type="bigint"/>
              <wlevs:property name="lastts" type="bigint"/>
           </wlevs:properties>
        </wlevs:event-type>
    </wlevs:event-type-repository>
    <wlevs:adapter id="InputAdapter" provider="csvgen">
         <wlevs:instance-property name="port" value="9061" />
         <wlevs:instance-property name="eventTypeName" value="InputEvent" />
         <wlevs:instance-property name="eventPropertyNames" value="id,obs,epoch" />
    </wlevs:adapter>
    <wlevs:channel id="InputChannel" event-type="InputEvent" >
        <wlevs:listener ref="helloworldProcessor"/>
        <wlevs:source ref="InputAdapter"/>
    </wlevs:channel>
    <wlevs:caching-system id="local-caching"/>
    <wlevs:cache id="lastMatchCache" name="lastMatchCache" key-properties="keyValue" value-type="CacheEvent">
       <wlevs:caching-system ref="local-caching"/>
    </wlevs:cache>
    
    <wlevs:processor id="helloworldProcessor">
       <wlevs:listener ref="CacheChannel"/>
       <wlevs:cache-source ref="lastMatchCache" />
    </wlevs:processor>
    <wlevs:channel id="CacheChannel" event-type="CacheEvent">
       <wlevs:listener ref="lastMatchCache"/>
    </wlevs:channel>
    <wlevs:channel id="OutputChannel" event-type="OutputEvent" advertise="true">
        <wlevs:listener>
            <bean class="com.bea.wlevs.example.helloworld.HelloWorldBean"/>
        </wlevs:listener>
        <wlevs:source ref="helloworldProcessor"/>
    </wlevs:channel>And the output that I get seems to match your requirement -
eventType=OutputEvent object=q1 v1.id=4143, v1.obsTotal=9, v1.elapsedTime=7228000, v1.firstts=1199311210000, v1.lastts=1199318438000
eventType=OutputEvent object=q1 v1.id=4143, v1.obsTotal=17, v1.elapsedTime=30935000, v1.firstts=1199318438000, v1.lastts=1199349373000
eventType=OutputEvent object=q1 v1.id=4143, v1.obsTotal=7, v1.elapsedTime=41436000, v1.firstts=1199349373000, v1.lastts=1199390809000
eventType=OutputEvent object=q1 v1.id=4143, v1.obsTotal=13, v1.elapsedTime=5904000, v1.firstts=1199390809000, v1.lastts=1199396713000Edited by: Anand Srinivasan on Sep 30, 2010 12:49 AM
Edited by: Anand Srinivasan on Sep 30, 2010 12:50 AM

Scripting help for image pop-up within a PDF

Hi,
I'm creating some sample eBooks in Acrobat Professional 9 (Mac), and I am stuck on one element that needs to be present. In the text within the PDF, there might be a phrase like, "See Fig. 1". Fig. 1 might not be on that page, however. I know how to create a link in a PDF that will navigate the reader to the page with the image referenced, but that's not good enough. I'd like to have this happen:
User clicks on "See Fig. 1".
Popup window appears with the referenced thumbnail image.
The image's popup window will stay on-screen until dismissed by the user.
Clicking on the popup thumbnail image calls up a full-size image.
The added wrinkles are many after we figure out the scripting for the above, but here are two major ones to be addressed right away:
1) The images are already part of the typeset document converted to PDF, so I am guessing the images would have to be bundled with the PDF somehow. Our users need to use the PDFs in an offline environment, so a call-out to a URL is out of the question.
2) The PDFs will be read on Macs, PCs, and possibly iPads. If a bundled, local file is called out when a user clicks on "See Fig. 1", how do I set a relative filepath in the call-out?
Thanks for any help you can provide.
Andrew

Here's a sample document that demonstrates some of what you're asking about: https://acrobat.com/#d=aP-2xL45pM57xVkT6d6fSg
The images are stored in buttons, so there is no concern about having to use external images. Note that this may not work on an iPad since most (all?) PDF viewers on the iPad do not have JavaScript support.

Help needed with pattern matching (analytics SQL ?)

Hello Forum Users,
I've got a curious problem on my hands that I am unable to think of an efficient way to code in Oracle SQL ....
(1) Background
Two tables :
MEASUREMENTS
ID NUMBER
M1 NUMBER
M2 NUMBER
M3 NUMBER
M4 NUMBER
INTERESTING_VALUES
ID NUMBER
M1 NUMBER
M2 NUMBER
M3 NUMBER
M4 NUMBER
(2) OUTPUT NEEDED
For each row in "MEASUREMENTS", count the number of matches in "INTERESTING_VALUES" (all rows). Please note that the matches might not be in the same column.... e.g. for a given row MEASUREMENTS M1 might match INTERESTING_VALUES M3.
I need to count 2,3 and 4 matches and provide the count values seperatley.
I am not interested in fuzzy matching (e.g. "greater than" or "less than"), the numbers only need to match exactly.
(You can use features up to 11g).
Thank you for your help.

yup here you go...
SQL> select * from measurement;
        ID         M1         M2         M3         M4
         1         30         40        110        120
         2         12         24        175        192
         3         22         35        147        181
SQL> select * from interesting;
        ID         M1         M2         M3         M4
         1         16        171         30        110
         2         40        171         30        110
         5        181        147         35         22
         4        175         12        192         24
         3        175         86        192         24
SQL> SELECT id,
2    SUM(
3    CASE
4      WHEN LEN=1
5      THEN 1
6      ELSE 0
7    END) repeat2,
8    SUM(
9    CASE
10      WHEN LEN=2
11      THEN 1
12      ELSE 0
13    END) repeat3,
14    SUM(
15    CASE
16      WHEN LEN=3
17      THEN 1
18      ELSE 0
19    END) repeat4
20     FROM
21    (SELECT id,
22      spath   ,
23      (LENGTH(spath)-LENGTH(REPLACE(LOWER(spath),',') ))/ LENGTH(',') LEN
24       FROM
25      (SELECT t.* ,
26        ltrim(sys_connect_by_path(cvalue,','),',') spath
27         FROM
28        (SELECT p.*,
29          row_number() over (partition BY id, id1 order by cvalue) rnum
30           FROM
31          (SELECT m.id,
32            m.m1      ,
33            m.m2      ,
34            m.m3      ,
35            m.m4      ,
36            (SELECT COUNT(*)
37               FROM interesting q
38              WHERE q.id=I.id
39            AND (q.m1   = column_value
40            OR q.m2     =column_value
41            OR q.m3     =column_value
42            OR m4       =column_value)
43            ) cnt              ,
44            column_value cvalue,
45            i.id id1           ,
46            i.m1 m11           ,
47            i.m2 m21           ,
48            i.m3 m31           ,
49            i.m4 m41
50             FROM measurement m                   ,
51            TABLE(sys.odcinumberlist(m1,m2,m3,m4)),
52            interesting I
53         ORDER BY id,
54            id1     ,
55            cvalue
56          ) p
57          WHERE cnt=1
58        ) t
59        WHERE level        >1
60        CONNECT BY prior id=id
61      AND prior id1        =id1
62      AND prior rnum      <=rnum-1
63        --start with rnum=1
64     ORDER BY id,
65        id1     ,
66        cvalue
67      )
68   ORDER BY 1,2
69    )
70 GROUP BY id
71 ORDER BY 1;
        ID    REPEAT2    REPEAT3    REPEAT4
         1          4          1          0
         2          9          5          1
         3          6          4          1I am posting the code without number formatting so that it gets easier for you to copy the code...
select id,sum(case when len=1 then 1 else 0 end) repeat2,sum(case when len=2 then 1 else 0 end) repeat3,sum(case when len=3 then 1 else 0 end) repeat4 from
(select id,spath,(LENGTH(spath)-LENGTH(REPLACE(LOWER(spath),',') ))/ LENGTH(',') len from
(select t.*
,ltrim(sys_connect_by_path(cvalue,','),',') spath
from (select p.*,row_number() over (partition by id, id1 order by cvalue) rnum from (SELECT m.id,
    m.m1      ,
    m.m2      ,
    m.m3      ,
    m.m4      ,
    (SELECT COUNT(*)
       FROM interesting q
      WHERE q.id=I.id
    AND (q.m1    = column_value
    OR q.m2     =column_value
    OR q.m3     =column_value
    OR m4       =column_value)
    ) cnt              ,
    column_value cvalue,
    i.id id1           ,
    i.m1 m11           ,
    i.m2 m21           ,
    i.m3 m31           ,
    i.m4 m41
    FROM measurement m                     ,
    TABLE(sys.odcinumberlist(m1,m2,m3,m4)),
    interesting I
order by id,id1,cvalue
) p where cnt=1
) t where level>1
connect by prior id=id and prior id1=id1 and prior rnum<=rnum-1
order by id,id1,cvalue) order by 1,2) group by id order by 1Ravi Kumar
Edited by: ravikumar.sv on Sep 15, 2009 3:54 PM

Help for Image detection in video & hardware issues

Hi,
I am trying to develop a web app that lets the user select shapes in a video(user can click only on detected shape if clicked elsewhere it should inform user to click only on the detected shape)
Then the location of the clicked position is determined & a robotic arm(at remote location) is moved to that position
I am a newbie on JMF
I do have a solution: load video on an applet(which can help in determining location of click but how would it know that the clicked position has the detected shape or not)
And here are my questions:
Is solution right?
Are there other solutions that i miss up?
What is the best language for hardware (microcontroller) programming here?
Is interaction between JMF & remote hardware possible & how?
Thank u all in advance

saurabh.nigam wrote:
Is solution right?The thing you're calling a "solution" doesn't appear to be a solution. Normally, solutions don't have questions embedded in them...
Are there other solutions that i miss up?Again, you're not being clear and you're nowhere near a solution...
What is the best language for hardware (microcontroller) programming here?That's not a question for this forum. This forum is for questions about the JMF API...
Is interaction between JMF & remote hardware possible & how?Yeah, you can always send commands via a socket using normal Java stuff. JMF isn't a special langauge or anything, it's just some Java classes that do media stuff...
In order to do what you're wanting to do, you'd have to implement a series of "not easy" things to implement.
1) You'd need to implement some sort of blob detection / extraction code that would take in an image and detect / extract the blobs in it.
2) You'd need to implement some sort of blob shape filter that would filter out all of the detected blobs that aren't a shape you're happy with
3) You'd need to embed your blob code into a JMF effect class so it could process video frames as they come through
4) You'd probably also want to augment the video stream with outlines or translucent colored regions to indicate to the user where the clickable blobs are...
5) You'd need to implement a socket-based control protocol for controlling the remote hardware.
All of that is possible with Java / JMF, but none of it is easy.

Need Help for Image Processing

Hi everyone,
my name is Davide and I am a new member. I apologize in advance for my English that It's not very good.
I have to do an exercise where I have to draw a SAR(Synthetic Aperture Radar) Image. The problem is that the input image is composed by a binary raster plus an XML parameters file. The XML file contains all the information to correctly read the image.
I downloaded Java Advancing Imaging Library but I've never worked with this image format and I don't know where to start.
Can anyone help me? How should i do with the binary raster file? How can i read it?
Thank you very much.

Hi everyone again,
I followed your advice and I studied the theory on raster data so I have made some step forward. Now I have a new problem that I will try to explain you in the better way. I was able to obtain from the binary raster data an array of float.
From the XML parameters file I know that the rows number are 1000, the columns number are 2000, the pixel type is complex float and that the real and imaginary parts are put side-by-side on a column basis. I created two array, one of the real parts and one of the imaginary parts:
for (int i=0;i < SIZE_X*SIZE_Y; i++){
   if(i%2 == 0){
      real[k] = bin_input;
if(Math.abs(real[k]) > max_value_re){
max_value_re = Math.abs(real[k]);
k++;
}else{
img[n] = bin_input[i];
if(Math.abs(img[n]) > max_value_im){
max_value_im = Math.abs(img[n]);
n++;
Then I create an array with the absolute value of complex number:for (int i=0;i < SIZE_X/2*SIZE_Y; i++){
mag[i] = (float) (Math.sqrt(Math.pow(real[i]/max_value_re,2) + Math.pow(img[i]/max_value_im,2)));
Now I have an array of float with values from 0 and 1. I tried to create a gray scale image because in the XML file it says that the number of channels is 1:BufferedImage theImage = new BufferedImage(SIZE_X/2, SIZE_Y, BufferedImage.TYPE_4BYTE_ABGR);
WritableRaster wr = (WritableRaster) theImage.getData();
wr.setPixels(0, 0, SIZE_X/2, SIZE_Y, mag);
File outputfile = new File("C:\\Sviluppo\\workspace\\Java\\SAR\\Milano.png");
try {
ImageIO.write(theImage, "png", outputfile);
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
The problem is that the image is all black as if the method is not able to interpret the float values and puts everything at zero.
First of all, do you think the process can be right or is there some logical/theorical error? If it's ok, How can I fix the problem?
Thank you.

Help with regex pattern matching.

Hi everyone
I am trying to write a regex that will extract each of the links from a piece of HTML code. The sample piece of HTML is as follows:
<td class="content" valign="top">
     
<script language="JavaScript"
src="http://chat.livechatinc.net/licence/1023687/script.cgi?lang=en&groups=0"></script>
<a href="makeReservation.html">Making a reservation</a><br/>
<a href="changeAccount.html">Changing my account</a><br/>
<a href="viewBooking.html">Viewing my bookings</a><br/>I am interested in extracting each link and the corrresponding text for that link into groups.
So far I have the following regex <td class="content" valign="top">.*?<a href="(.*?)">(.*?)</a><br>However this regex only matches the first line in the block of links, but I need to match each line in the block of links.
Any ideas? Any suggestions are appeciated as always.
Thanks.

Hi sabre,
thanks for the reply.
I am already using a while loop with matcher.find(), but it still only returns the first link based on my regex.
the code is as follows.
private static final Pattern MENU_ITEM_PATTERN = compilePattern("<td class=\"content\" valign=\"top\">.*?<a href=\"(.*?)\">(.*?)</a><br>");
private LinkedHashMap<String,String> findHelpLinks(String body) {
    LinkedHashMap<String, String> helpLinks = new LinkedHashMap<String,String>();
    String link;
    String linkText;
      Matcher matcher = MENU_ITEM_PATTERN.matcher(body);
      while(matcher.find()){
        link = matcher.group(1);
        linkText = matcher.group(2);
        if(link != null && linkText != null){
          helpLinks.put(link,linkText);
    return helpLinks;
private static Pattern compilePattern(String pattern) {
    return Pattern.compile(pattern, Pattern.DOTALL + Pattern.MULTILINE
        + Pattern.CASE_INSENSITIVE);
}Any ideas?

Help for Image Distortion Anyone?

Is there an easy way to get one line of distortion through text like shown in the picture?
or maybe a similar effect?
Thank you for any suggestions.

Good day!
Depending on the intended use of the text it might be advisable to use vector data.
If you want to use pixel data there are several options.
I would recommend using the Filter Displace as a Smart Filter with a custom Displacement Map.
Regards,
Pfaffenbichler

Need help for image flipping

How can i flip an image without using g.drawImage method
bcos i have already used this method in
public void paint(Graphics g)
and i can not use this method again it print out errors..
i want to use anthoer g.drawimage

here's a short program that loads an image, flips it, and shows it on screen:import java.awt.geom.AffineTransform;
import java.awt.image.AffineTransformOp;
import java.awt.image.BufferedImage;
import java.io.File;
import javax.imageio.ImageIO;
import javax.swing.ImageIcon;
import javax.swing.JFrame;
import javax.swing.JLabel;
class Flip {
    public static void main(String[] args) throws Exception {
     // Load jpg file to a BufferedImage.
     File f = new File("/home/jsalonen/Desktop/small.jpg");
     BufferedImage bi = ImageIO.read(f);
     // Flip that image.
     BufferedImage flipped = new BufferedImage(
                   bi.getHeight(),
                   bi.getWidth(),
                   BufferedImage.TYPE_INT_RGB);
     AffineTransform trans = new AffineTransform(0, 1, 1, 0, 0, 0);
     AffineTransformOp op = new AffineTransformOp(trans,
                   AffineTransformOp.TYPE_NEAREST_NEIGHBOR);
     op.filter(bi, flipped);
     // Show it on screen.
     JFrame jf = new JFrame("Flip");
     jf.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
     JLabel showLabel = new JLabel(new ImageIcon(flipped));
     jf.getContentPane().add(showLabel);
     jf.pack();
     jf.show();
}

How can I load pattern matching images into memory?

I discovered this problem by accident when our network went down. I could not run my VI because the .png file used for pattern matching was on a network drive. My concern it the amount of time that is required to read a network file. Is there a way to load the file into memory when the VI is opened?

Brian,
Thank you for contacting National Instruments. For most pattern matching programs, the pattern file should only be read from file once and is then copy to a buffer on the local machine. If you examine your code, or an example program for pattern matching, you should see at least two IMAQ Create VI called somewhere near the front of your code. This VI basically creates a memory location and most likely will be called once for your pattern image and once for the image you are searching.
Unless you are specifically calling a File Dialog VI where you are given a dialog box to open a file or have hard coded the file path so that it is read each iteration of your code, then your pattern file should only need to be called at the beginning of your application, th
us causing only one file read over the network for that image. Therefore your program most likely already loads the image in memory once it is read and should not be accessing the network constantly.
Again, I would recommend taking a look at your code to make sure you are not causing a file access every time and then you should be ready to go. Just in case you do have the network go down, I would recommend keeping a copy of the image locally and, if you are feeling ambitious, you can programmatically have your program read the file locally if the network file returns an error.
Good luck with your application and let us know if we can be of any further assistance.
Regards,
Michael
Applications Engineer
National Instruments

Color pattern matching is very slow

Hi
I tried this code creating one vi application.
After the testing with USB webcam I have realized that the color pattern matching is very slow. How to increase the speed and to work smoothly in real time.
Thank you

Hello tiho,
the color pattern matching is not as fast as 8-bit matching, but should still be fast.
For example, I am attaching a VI for color pattern matching where you load the image, create the template and search do the matching.
In my example I tried color pattern matching on color image of size 4288x2848 pixels and the matching is performed in ~140 ms (~7Hz). So, for a smaller image, I think the real-time processing is quite achievable (I consider real-time 20 Hz or more). The only problem is the template learning, which in my case takes around 10 seconds. But you should learn the template only once in the initialization stage.
Best regards,
K
https://decibel.ni.com/content/blogs/kl3m3n
"Kudos: Users may give one another Kudos on the forums for posts that they found particularly helpful or insightful."
Attachments:
color matching.zip ‏49 KB

Pattern matching program

Hello,
I am creating a program that does pattern matching on a video. I want the program to be able to open a template to be used for the pattern matching template and to allow the user to select a template from an image in Labview and then use this template as the pattern matching template. I have code that does all of this. The part that allows a user to select a template in Labview is a sample program that comes with Labview. I just do not know how to incorporate it. Thanks.
Solved!
Go to Solution.

Hi frog,
Here in attached vi, Ckeck it out your problem solved or not. If not, post the video file also. will help you..
Not to mention: Your vi is so big. You have to really optimize your vi.
Sasi.
Certified LabVIEW Associate Developer
If you can DREAM it, You can DO it - Walt Disney
Attachments:
Pattern Tracking 2.vi ‏135 KB

Gige + pattern matching

Hi all,
I am new to Labview image processing .
i need to achive pattern match. from that i need to find fiducail location . but here i am facing some problems like .
1. aquiring image is corrupted like black lines (i am using GigE camera 5MP 2/3" ) .
2. here i am doing pattern match by live grab method ( like continous aquring) . so that my CPU uasge is reachs nearlly 90 %
3. my task is to find fiducail using pattern match from that i need to assign ROI for main pattern (is it possible to snap image -> pattern match - > result)
i have attached my vi code and image ..
Attachments:
6.png ‏75 KB
PM_TEST.vi ‏121 KB

Hi please learn basics and then apply over...
1)What are the black lines in image? you didn't provide any screenshot/corrupted image.
-Are you getting images properly NI MAX?
-Did you allow jumbo packets ?
2)Please understand what is grab and snap by using built-in examples provides. And your cpu usage is not because of continuous acquisition.
-I cannot comment on the cpu usage now, but sure it's not because of gab.
-Use producer-consumer pattern to do parallel acquisition and processing.
3)Once you understand the difference between grab and snap, this you'll understand automatically.
Thanks
uday,
Please Mark the solution as accepted if your problem is solved and help author by clicking on kudoes
Certified LabVIEW Associate Developer (CLAD) Using LV13

How can I use SQL to search for a pattern within a field?

Hello, Frank, Solomon, ect
I am now faced with this particular scenario, I've got the SQL to search through a field to find text within the field, but I have to know what it is before it can look for it.
What I have to do is this:
Search through a field, for a pattern, and I won't know what the data is I am looking for. Can this be done in SQL?
For instance, Here is my SQL this far, I was helped allot in order to get to this point.
select table_name,
       column_name,
       :search_string search_string,
       result
from (select column_name,
               table_name,
               'ora:view("' || table_name || '")/ROW/' || column_name || '[ora:contains(text(),"%' || :search_string || '%") > 0]' str
          from cols
         where table_name in ('TABLE1', 'TABLE2')),
       xmltable (str columns result varchar2(10) path '.')
When you execute the above SQL, you have to pass in a value. What I really need is to alter the above SQL, to make it search for a pattern that exist's within the text of the field itself.
Like for instance, lets say the pattern I am looking for is this" xx-xxxxx-xxxx" and it's somewhere in a field.
I need to alter this SQL to take this pattern and search through all the schemas and tables to look for this pattern match.
Can be done?

When you use something dynamically within a function or procedure, roles do not apply and privileges must be granted directly. So, you need to grant select on dba_tab_cols directly. If you want to do pattern matching then you should use regular expressions. The following example grants the proper privileges and uses regexp_instr to find all values containing the pattern xxx-xxxx-xxxx, where /S is used for any non-space character. I limited the tables in order to save time and output for the test, but you can eliminate that where clause.
SYS@orcl> CREATE USER test IDENTIFIED BY test
2 /
User created.
SYS@orcl> ALTER USER test QUOTA UNLIMITED ON USERS
2 /
User altered.
SYS@orcl> GRANT CREATE SESSION, CREATE TABLE TO test
2 /
Grant succeeded.
SYS@orcl> GRANT SELECT ON dba_tab_cols TO test
2 /
Grant succeeded.
SYS@orcl> CONNECT test/test
Connected.
TEST@orcl> SET LINESIZE 90
TEST@orcl> CREATE TABLE table1
2    (tab1_col1 VARCHAR2(60))
3 /
Table created.
TEST@orcl> INSERT ALL
2 INTO table1 (tab1_col1) VALUES ('xxx-xxxx-xxxx')
3 INTO table1 (tab1_col1) VALUES ('matching abc-defg-hijk data')
4 INTO table1 (tab1_col1) VALUES ('other data')
5 SELECT * FROM DUAL
6 /
3 rows created.
TEST@orcl> CREATE TABLE table2
2    (tab2_col2 VARCHAR2(30))
3 /
Table created.
TEST@orcl> INSERT ALL
2 INTO table2 (tab2_col2) VALUES ('this BCD-EFGH-IJKL too')
3 INTO table2 (tab2_col2) VALUES ('something else')
4 SELECT * FROM DUAL
5 /
2 rows created.
TEST@orcl> VAR search_string VARCHAR2(24)
TEST@orcl> EXEC :search_string := '\S\S\S-\S\S\S\S-\S\S\S\S'
PL/SQL procedure successfully completed.
TEST@orcl> COLUMN "Searchword"     FORMAT A24
TEST@orcl> COLUMN "Table"     FORMAT A6
TEST@orcl> COLUMN "Column/Value" FORMAT A50
TEST@orcl> SELECT DISTINCT SUBSTR (:search_string, 1, 24) "Searchword",
2               SUBSTR (table_name, 1, 14) "Table",
3               SUBSTR (t.column_value.getstringval (), 1, 50) "Column/Value"
4 FROM   dba_tab_cols,
5          TABLE
6            (XMLSEQUENCE
7           (DBMS_XMLGEN.GETXMLTYPE
8              ( 'SELECT ' || column_name ||
9               ' FROM ' || table_name ||
10               ' WHERE REGEXP_INSTR
11                     (UPPER (' || column_name || '),''' ||
12                  UPPER (:search_string) || ''') > 0'
13              ).extract ('ROWSET/ROW/*'))) t
14 WHERE table_name IN ('TABLE1', 'TABLE2')
15 ORDER BY "Table"
16 /
Searchword               Table Column/Value
\S\S\S-\S\S\S\S-\S\S\S\S TABLE1 <TAB1_COL1>matching abc-defg-hijk data</TAB1_COL1>
\S\S\S-\S\S\S\S-\S\S\S\S TABLE1 <TAB1_COL1>xxx-xxxx-xxxx</TAB1_COL1>
\S\S\S-\S\S\S\S-\S\S\S\S TABLE2 <TAB2_COL2>this BCD-EFGH-IJKL too</TAB2_COL2>
3 rows selected.

Pattern-Matching Syntax in NPS

Hi All,
I am looking for help in creating pattern matching syntax for 8 IP addresses in Network Policies. Want to add condition of NAS IPv4 Address in a single policy. Ip Addresses are:
10.110.7.38
10.110.7.39
10.110.23.38
10.110.23.39
10.110.43.36
10.110.43.37
192.194.135.152
192.194.135.153
all ip addresses in a single NAS condition. Please assist me as pattern matching syntax guide is confusing for me.

Hi,
The guide at https://technet.microsoft.com/en-us/library/dd197583(v=ws.10).aspx indicates:
x|y
Matches either x or y.
So you can try:
10.110.7.38|10.110.7.39|10.110.23.38|10.110.23.39|10.110.43.36|10.110.43.37|192.194.135.152|192.194.135.153
Let me know if this doesn't work.
-Greg

Help for image pattern matching

Similar Messages

Maybe you are looking for