Hi ABBYY community!
Please help me to solve my problem:
I have several flexible forms, on which the sizes and locations of field may vary from form to form.
Data(numeric value, must be 12 digit number) which I need to extract can be in
- conjoined frames for characters
or
- frames with "combs"
IN FLS 10 I've created element (type "object collection") and in Abbyy FlexiCapture 10 I use script to clean recognized value from different garbage and property "conjoined frames for characters" in tab "recognition"
Result:
1) Data from conjoined frames is what I want
2) Data from forms with frames with "combs" is less then 12 digit number(image property: size 1616*2296, 200 dpi)
How to recognize words in special frames?
Please help me to solve my problem:
I have several flexible forms, on which the sizes and locations of field may vary from form to form.
Data(numeric value, must be 12 digit number) which I need to extract can be in
- conjoined frames for characters
or
- frames with "combs"
IN FLS 10 I've created element (type "object collection") and in Abbyy FlexiCapture 10 I use script to clean recognized value from different garbage and property "conjoined frames for characters" in tab "recognition"
Result:
1) Data from conjoined frames is what I want
2) Data from forms with frames with "combs" is less then 12 digit number(image property: size 1616*2296, 200 dpi)
How to recognize words in special frames?
コメント
7件のコメント
Could you please provide some sample images and screenshots of your current results for better understanding?
Here it is
Could you please try sending images again, it seems like something went wrong and they weren't properly added.
It seems like the issue with attachments is still present, could you please upload these images to dropbox or googledrive?
https://drive.google.com/file/d/0B25H0vMsyN7CWDc5aGkzUXBIM28/view?usp=sharing
https://drive.google.com/file/d/0B25H0vMsyN7CalQxQ0tfR1lQT3M/view?usp=sharing
https://drive.google.com/file/d/0B25H0vMsyN7CUm02Q3U5a08tRlk/view?usp=sharing
https://drive.google.com/file/d/0B25H0vMsyN7CeU4xWG0waFZpUTg/view?usp=sharing
In order to find optimal solution you need to try several diffrent approaches:
1.Try using diffrent recognitions options (frames, combos etc)
2.Or use standard recognitions options and try to remove garbage of specified size 10-20 dots for example, this might help removing box contours and improve recognition
3. Adjust your flexi layout:
- use a character string to search for numerical symbols, you can specify character string to capture only numerical symbols and have a certain limit to max gap between characters in order to capture the whole string. Then you can use this string as source for your textblock
4.Another way to adjust flexilayout is:
- create a region element that would contain all numerical symbols
- capture each numerical symbol invidualy using a repeatable group with search area limited by previous region element
- create a resulting region element with search area equal to combined areas of all repeatable groups' instances
This way you might be able to avoid capturing garbage
サインインしてコメントを残してください。