| ABBYY Mobile Imaging SDK 1.0 |
Go to: Contents | Programmer's Guide
Defect Detectors
Note. It is assumed that, in addition to general considerations on the use of detectors, the type of images, where they will be used, is of great importance. Therefore, the threshold values for the motion blur and defocus detector, that would allow to differentiate bad images from good ones, are not provided in the recommendations.
Currently, there are three types of quality detectors:
The detector returns the following values:
You can use one of these parameters to assess whether their is sufficient light or not. The signal to noise ratio is intended for customization, while the binary value is set during the detection.
The detector returns the following values:
Before looking for defects, the entire image is divided into blocks, and each block is analyzed to determine whether it has any text or not. All subsequent information relates to the area of the image made up of blocks containing the text.
The detector returns the following values:
To ensure high-quality recognition, glare and noise detectors should return "false" and the defocus detector should return close to "zero" value for the size of the blurred and defocused areas. It is assumed that, for typical images, the threshold value of the latter parameters is defined by the user as it may differ in various types of images.