Atalasoft MobileImage API Reference
Properties | List of all members
kfxKEDDocumentBaseDetectionSettings Class Reference

Base class for detector settings. More...

#import <kfxKEDDocumentBaseDetectionSettings.h>

Inheritance diagram for kfxKEDDocumentBaseDetectionSettings:
kfxKEDCheckDetectionSettings kfxKEDDocumentDetectionSettings

Properties

CGPoint targetFrameCenter
 The center point within the image of the target frame. More...
 
CGFloat targetFramePaddingPercent
 The minimum padding in percent length to be maintained between the edges of the target frame and image. More...
 
CGFloat targetFrameAspectRatio
 The aspect ratio of the rectangular frame that should match the document being searched. More...
 
CGFloat zoomMinFillFraction
 The minimum fill area as a fraction of the target frame. More...
 
CGFloat zoomMaxFillFraction
 The maximum fill area as a fraction of the target frame. More...
 
CGFloat turnSkewAngleTolerance
 The maximum angle the document can be rotated off from the target frame. More...
 
CGFloat verticalMovementTolerance
 The maximum distance the document can be shifted above or below the target frame. More...
 
CGFloat horizontalMovementTolerance
 The maximum distance the document can be shifted left of right of the target frame. More...
 

Detailed Description

Base class for detector settings.

The primary purpose of the settings object is to define properties of a rectangular target frame. The target frame is a rectangular area confined within the bounds of the image, and represents the the ideal area that should be covered by a document. Results returned from the document detector will provide adjustment guidance relative to the target frame.

A target frame is constructed from a center point, an aspect ratio, and a minimum edge padding. The size of the target frame is the largest possible rectangle that can be made without violating any of these constraints.

Additional settings properties set tolerances on the guidance available from a result object.

Property Documentation

- (CGFloat) horizontalMovementTolerance
readwritenonatomicassign

The maximum distance the document can be shifted left of right of the target frame.

The horizontal movement tolerance puts an upper bound on how far a document can be shifted left of right of the target frame and still be considered acceptable. The tolerance is specified as a fraction of the target width, and the comparison is made between the center of the target frame and the center of the detected document bounds.

The default tolerance value is 0.15.

- (CGFloat) targetFrameAspectRatio
readwritenonatomicassign

The aspect ratio of the rectangular frame that should match the document being searched.

The aspect ratio is presented in the format of longEdge/shortEdge, and is used to construct the target frame. There are two common check aspect ratios:

Personal: (6.0 / 2.75) = 2.18 Business: (9.25 / 3.67) = 2.52

Aspect ratio between 0 and 1 or negative is not recommended. It can result in erroneous detection of target frame. If the targetFrameAspectRatio is set to 0, then it will be automatically reset to an image aspect ratio.

The default value is the personal check ratio, 2.18.

- (CGPoint) targetFrameCenter
readwritenonatomicassign

The center point within the image of the target frame.

The center point is any point within the image that the target frame will be constructed around and centered over. This value will be clamped the valid area of the image being searched, and by default is set to the center of the image.

- (CGFloat) targetFramePaddingPercent
readwritenonatomicassign

The minimum padding in percent length to be maintained between the edges of the target frame and image.

Given a center point and aspect ratio, a rectangular frame will be constructed to cover as much of the image as possible without any the distances between a frame edge and image edge falling below the padding distance, calculated as percent of the perpendicular image edge length. Values will be clamped to the range [1-50]. The default value is 5.

For example:

  • Given detection settings with a frame center at the center of the image, and a personal check aspect ratio, the constructed frame will fill the width of the image, leaving a space on the left and right equal to the padding percent of the image width.
  • Given detection settings with a frame center shifted to the left of the image center, the gap on the right side of the constructed frame will exceed the gap on the left.
See also
targetFrameCenter
targetFrameAspectRatio
- (CGFloat) turnSkewAngleTolerance
readwritenonatomicassign

The maximum angle the document can be rotated off from the target frame.

The skew angle tolerance puts an upper bound on how far a document can be rotated clockwise or counter-clockwise relative to the target frame and still be considered acceptable. Angles of rotation greater than the tolerance will result in guidance to turn the document. The default value is 10, specified in degrees.

- (CGFloat) verticalMovementTolerance
readwritenonatomicassign

The maximum distance the document can be shifted above or below the target frame.

The vertical movement tolerance puts an upper bound on how far a document can be shifted above or below the target frame and still be considered acceptable. The tolerance is specified as a fraction of the target height, and the comparison is made between the center of the target frame and the center of the detected document bounds.

The default tolerance value is 0.15.

- (CGFloat) zoomMaxFillFraction
readwritenonatomicassign

The maximum fill area as a fraction of the target frame.

The maximum fill fraction puts an upper bound on what is considered an acceptable zoom level. If the area of a detected document is greater then the area of the target frame multiplied by the minimum fill fraction, guidance will suggest zooming out from the document. The default value is 1.5.

- (CGFloat) zoomMinFillFraction
readwritenonatomicassign

The minimum fill area as a fraction of the target frame.

The minimum fill fraction puts a lower bound on what is considered an acceptable zoom level. If the area of a detected document is less then the area of the target frame multiplied by the minimum fill fraction, guidance will suggest zooming in on the document. The default value is 0.65.


The documentation for this class was generated from the following file:
Untitled Document © 2016 Atalasoft, Inc., 116 Pleasant St, Suite 321, Easthampton, MA 01027, U.S.A. All rights reserved. Use is subject to license terms.