Image processing allows for manipulation of scanned images either to improve quality of the image as it is imported or to gather data from the image itself. Pixel Supports a number of image processing filters for color and binary (black and white) image.
Filters
Although you can use all filters for both color and binary images, adding the Threshold filter to convert an image to black and white may improve the results if processed prior to another filter, such as Dilation. The order in which the filters appear in the Selected filters pane determines at what point during the image processing Perceptive Content performs the filter action. When you build your Scanning Profile as part of building your Capture Profile, the following filter options are available in the Image Processing tab when you choose Pixel as your device.
- Barcode Detection
- The Barcode Detection filter can be used to detect and decode barcodes on scanned image. the filter can detect many different types of barcodes, and there is no limit to the number of barcodes it can detect on a page.
- Black Overscan Removal
- The Black Overscan Removal filter can be used to remove the black area around an image when a page is scanned using an overscan option. Unlike the Border Removal filter, which inverts the black pixels and makes them white, and the Crop filter which removes white space from around the edge of the image, the Black Overscan removal filter removes the overscan area from your image. This filter reduces the physical size of the scanned image and the image file size by eliminating the black border generated by scanners with black backgrounds. For best results, deskew images with black fill before applying this filter. The Black Overscan removal filter has no configurable settings.
- Blank Page Detection
- The Blank Page Detection filter can be used to detect blank pages. Use the Sensitivity property to specify noise tolerance during detection.
- Border Removal
- The Border Removal filter can be used to remove black edges that sometimes appear around the image during scanning or photo-copying. The Border Removal filter looks for black around the edge of an image (borders) and changes it to white. Borders must be all black. If the border contains white, the filter will not remove it. The Border Removal filter has no configurable settings.
- Color Dropout
- The Color Dropout filter takes a color image and transforms specified colors into other colors. The filter maintains a list of color mappings that it should make when the filter is run. The Color Dropout Properties dialog allows the user to manage a list of colors that will be transformed.
- Crop
- The Crop filter can be used to set image margins and to remove white space around the edge of the image. You can set different values for the Top, Left, Bottom, and Right properties, or you can use the Force Symmetry property to lock the top/bottom and left/right margins to the same values.
- Deskew
- The Deskew filter can be used to straighten images that show a slant from its correct orientation. Skewing can occur if the original document was skewed when it was fed into the scanner, fax machine, or photocopier. The Deskew filter examines the image and determines the skew angle. The skew angle is measured between the actual edge of the image data and the horizontal or vertical axis. The image data is then rotated to correct the skew angle. Deskewing an image makes the image contents more legible and can improve OCR results.
- Dilation
- The Dilation filter can be used to expand the area of black objects in an image. Using this filter may improve image quality and the legibility of text. Performing dilation can increase the file size.
- Erosion
- The Erosion filter can be used to trim the area of black objects. Using this filter will reduce the file size, but you could lose a lot of detail from the image content.
- Halftone Removal
- The Halftone Removal filter can be used to remove a background from an image or a graphics object on the image. The Halftone Removal filter has no configurable settings.
- Hole Removal
- The Hole Removal filter can be used to remove punch holes from the edges of the image. The algorithm searches for objects which look like punch holes around the edges of the image. Punch hole-like graphics objects in other areas of the image will not be removed. The Hole Removal filter has no configurable settings.
- Invert Image
- The Invert filter can be used to invert the image to its negative equivalent. The filter actually modifies the image data itself, changing white pixels to black and black pixels to white. The Invert filter has no configurable settings.
- Line Removal
- The Line Removal filter can be used to remove lines from or to reconstruct lines on a form-based image. Removing lines can reduce the image file size and improve OCR results.
- Noise Removal
- The Noise Removal filter returns or sets the maximum area percentage an object can take up within the area defined by the maximum noise width and height in order for the object to be considered noise.
- Patchcode Detection
- The Patchcode Detection filter can be used to detect and decode patchcodes on scanned images. The Pathcode Detection filter has no configurable settings.
- Rotation
- The Rotation filter can be used to perform image rotation in 90 degrees increments. It can rotate the image 90 degrees clockwise and counterclockwise. it can also rotate it 180 degrees as well as perform mirroring, which flips the page across the vertical axis so that it appears to be the mirror image of the original.
- Scaling
- The Scaling filter can be used to resize an image while preserving the original aspect ratio. After you specify the desired width and height after scaling, the image area is resized so that it fits within those boundaries, while maintaining the aspect ratio. If either the desired height or width value is larger than the scaled image area, then the image area is centered along that dimension and white margins added to both sides.
- Skeleton
- The Skeleton filter can be used to reduce black objects in an image to one pixel thick skeletonized versions. Using this filter can reduce file size, but it can seriously distort the image. It should be used with caution and is usually only appropriate with certain types of OCR engines. The Skeleton filter has no configurable settings.
- Smoothing
- The Smoothing filter can be used to remove bumps and spurs that appear on text characters or graphics objects in the image. The filter looks for any pixel that is surrounded by five or six other connected pixels of the opposite color, and then inverts that center pixel based on the filter configuration. Smoothing is a good way to improve legibility or to reduce the file size without losing a lot of image detail.
- Threshold
- The Threshold filter can be used to convert a 24-bit color image to a binary image. All of the pixels in a color image that are darker than the threshold specified with the Brightness and Contrast properties are converted to black and all of the pixels that are lighter than the threshold are converted to white.