Hi, i want to perform an OCR process for the selection text in WebAnnotationViewer1 using RectangleData.
How can i do it? I'm currently test out the following code during OnAnnotationRightClicked event but it return empty string to me.

WebAnnotationViewer1.RemoteInvoke("GetWordAtPoint", new Array(x, y));

Any sample for the Web OCR using RectangleData?
Thanks.

Take a look at this application we made

http://www.atalasoft.com/cs/blogs/31appsin31days/archive/2008/05/22/clickword-online-ocr-by-click.aspx

Full source is provided.

Hi, i already downloaded that program.

Question 1
For the function below, is that 'alertWord' will called first, then follow by 'GetWordAtPoint'? For the x, y, is it the point i clicked on? And how if i wan to pass in the rectangle point?

   var go = function() {
        WebAnnotationViewer1.RemoteInvoked = alertWord;
        WebAnnotationViewer1.RemoteInvoke("GetWordAtPoint", new Array(x, y));
   }


Question 2
In GetWordAtPoint function, may i know what is w.Bounds.Contains(new Point(x, y)) doing? And i not understand why everytime i will get the same result for r value for OcrTextRegion, eventhough the image i loaded into WebAnnotationViewer1 is different. What i need to check out to solve this issues?

        [RemoteInvokable]
        public string GetWordAtPoint(int x, int y)
        {
            OcrPage p = (OcrPage)Application["ocrPage"];

            string val = "";

            foreach (OcrTextRegion r in p.Regions)
                foreach (OcrLine l in r.Lines)
                    foreach (OcrWord w in l.Words)
                        if (w.Bounds.Contains(new Point(x, y)))
                            val = w.Text;

            // This is not the best way to escape single quotes.
            // It should be much more robust than this.
            return val.Replace("'", "~~");
        }

Question 3
When should i call up this two events? Can i called it during OpenImageCallBack() function? Or must called during OnPageLoad() function?

WebAnnotationViewer1.RemoteInvoked = showPanel;
WebAnnotationViewer1.RemoteInvoke("OcrImage", new Array());

 

function OpenImageCallBack(){
 WebAnnotationViewer1.RemoteInvoked = function(){};
 WebThumbnailViewer1.OpenUrl(WebAnnotationViewer1.getReturnValue());
 Label_FileName.innerHTML = WebThumbnailViewer1.getUrl();
 WebAnnotationViewer1.RemoteInvoked = showPanel;
 WebAnnotationViewer1.RemoteInvoke("OcrImage", new Array());
}

Question 4
The sample source code not support in IE7? I test out the same copy of source code, it run fine in IE6, but in IE7 it return empty string when i do the click.

 Thanks.

Greetings!

To answer your questions...

1.  A function that is set to the RemoteInvoked (note the past tense) property of the WebImageViewer or WebAnnotationViewer is run AFTER the RemoteInvoke Call.   So, in this example, SetWordAtPoint runs, then alertWord is called when the server-side processing is finished.

2.  w.Bounds.Contains(new Point(x, y))

Each OcrWord w that is return contains a bounding rectangle.  If the point you clicked is inside that rectangle, the OcrWord is returned as the correct word.

This demo was not set up for selection rectangles -- only single clicks.  The OcrTextRegion will always be the entire page in this example, as the initial Recognize was not done with any sort of region rectangle.

3.  The OcrImage function should be called in the OnPageLoad function.  It is an initializing function you only want to call once.

4.  I am not having issues with IE7.  However, I've attached a different demo for you -- one that incorporates selecting the text you want recognized with a rectangle, and only performing OCR on the region specified.