Image enhancement options in Enhance Image action

Improve the quality of an image that you have uploaded in the Enhance Image action by using additional options available.

Option Use
Grayscale Convert the color pixels in an image to grayscale.

Recommended: Convert color images to grayscale to improve the OCR extraction accuracy.
Blur

Smoothen out areas that are pixelated and improve image quality.

Recommended values: Any odd number (such as 3, 7, and 11), as long as it is less than the pixel size of the uploaded image.
Sharpen

Improve clarity by removing distorted and blurred aspects of the original image.

Consider a scenario where you have an image file called ImageA. When you apply Gaussian blur = 5 to this image, a blurred version of this image is created called ImageB. You can then use the following Sharpen parameters:

  • Alpha: 3
  • Beta: -1
  • Gamma: 0

The following formula is applied for output:

Output = ImageA*Alpha + ImageB*Beta + Gamma

Where:

  • ImageA: First input array (Input image)
  • Alpha: Weight of the first array elements
  • ImageB: Second input array image with the same size and channel number as ImageA
  • Beta: Weight of the second array elements
  • Gamma: Scalar added to each sum
Note: The principal use of the preceding equation is to sharpen an image by removing the blurriness from the original image.

Recommended values:

  • Gaussian blur: 0 to 10
  • Alpha: 1 to 15
  • Beta: -0.5 to -5
  • Gamma: 0 to 50
Denoise

Smoothen out pixels that are not grayscale to remove distortion or speckling from the image.

Recommended value: 21. However, it can be any odd number as long as it is less than the pixel size of the uploaded image.
Contrast

Increase or decrease the contrast of the pixels. Values can range from 0 (no change) to 2 (maximum contrast).

Recommended values: From 0.8 through 2.
Brightness Increase or decrease the brightness level of all the pixels in the image. Values can range from -100 (minimum contrast) to 100 (maximum contrast), with 0 representing no change.

Recommended values: From 0 through 100
Thresholding A form of image segmentation. Thresholding is an image processing method that creates a bi-tonal (binary) image by setting a threshold value on the pixel intensity of the original image. In image processing, thresholding is typically used to separate objects or foreground pixels from background. For example, the following image has been processed using Thresholding value as 200 and 100 respectively.

thresholding value

Recommended value: From 0 through 255
Remove RGB Provide upper and lower bound values to pixels in a color image. Pixels in a color image are represented by three primary color channels: red, green, and blue (RGB). The output of this effect is a binary image with pixels that are black (0) or white (255). Consider a scenario where you want to transform high-intensity red pixels in the range of 200 to 255 to white (255). You can set the lower bound for red to (200,0,0) and upper bound to (255,0,0).