Please Help me

Hi

I am sending you an Image.

There are so many small images are embedded in a single image.

Now I want to identify all these images and store them in a database with name that is a text on the small images in the sent image.

I want to store these small images in this single image as separate images and tag the name.

Is it possible to identify these images and save separately using Lead tools OCR Document Imaging Tool?

If so please guide me how to proceed further or send me some examples.

Thanks in advance.

With Regards

Hemant BK

You did not have any files attached to your post.  Please attach it so I can better understand your problem.

NOTE: If you are attaching a project or file to the forums...

Hi

I am sending the Image as ZIP File and RAR File , my file size is 4.66mb

still when I post it , it do not get posted .

What could be the problem?

Thanks in advance

With Regards

 Hemant BK

 

 

Hi

Again I am sending the image.

Thanks in advance

with regards

 Hemant BK

 

The OCR engine will not be able to separate the images or aid in separating them because it deals with finding text. 

If you want to separate the small images (the photo IDs I presume) you will need to make a search algorithm that can detect where the IDs are located, and then crop them into their own images.  Unfortunately, LEAD does not have anything that can do this, since all of the IDs are different. 

My first idea would be to search by scanning horizontally and/or vertically and find the thick, black frame around the ID photos.  The frame is pretty distinguishable and looks to be the same size on each ID, so you should be able to make some pretty good guesses as to what is the frame and what isn't based on the DPI, number of pixels the same color (or color range) in a row.

Probably your biggest difficulty in this kind of algorithm is the fact that your image is grayscale, so the frame will not always be perfectly black.  Second biggest will be any ID pictures that are dark.

Hi

Thanks for your suggestions.

Shall I get some more suggestions from you?

My requirement is this.

I have big image (the sent one) which comprises some small sized images and text above indicating the name of the image.

I want to save these small images as separate images with the name which is above in the sent image.

There are two possibilities.

1. While scanning only I should scan these images as separate images by defining the regions in a template (scanning template).

Then I should define the regions and read these texts on the images using OCR / ICR tools.

Finally associate these images with the names read as text (using any of these ICR or OCR) and save them in the database.

2. If not in scanning then I should process this image. ie Define regions and separate these images and again read the texts on the image using OCR OR ICR readers and associate these names with the image and store it in the database.

In short the requirement can be divided into three

1. Extracting the Images. (Small images)

2. Reading the text on the images by defining the regions.

3. Associate these images with the names read and store them in the database.

The first one is either carried out while scanning the image by defining a template for scanning where I should be able to define regions for these images or after scanning the image I should process the image to extract the small images.

According to you which one easier?

Shall I save these images as different images while scanning by defining a template (where I should be able to define regions.)? If so how should I proceed? Does a Lead tool support this type of scanning?

Can you suggest me any Image processing tools which can help me to separate these images?

Is it possible to read these texts above the images using lead tools OCX?

Please throw some light on this issue.

 

With Regards

 Hemant BK

 

 

 

 

 

 

 

 

 

 

   

 

 

I would think that using your scanner to "cut" your large document up would be easiest.  That way you only have one photo ID and one set of information to worry about.  However, not all scanners will have this ability.  You will need to use the ICAP_FRAMES capability and the TW_IMAGELAYOUT structure.  For more information on these and how they are used, you will need to check the TWAIN 1.9 documentation (http://twain.org/downloads.shtm).  Within LEADTOOLS, you will be able to access the scanner by using the GetCapability and SetCapability methods.

There is no OCX interface for OCR, but you can use the OCR COM interface.  The LEADRaster.Bitmap property should be interchangeable with the LEAD Main Control's Bitmap property.