Automation Anywhere ドキュメントを読んで確認する

Automation 360

コンテンツを閉じる

コンテンツ

コンテンツを開く

Image enhancement options in EnhanceImage action

  • 更新済み: 2022/01/18
    • Automation 360 v.x
    • ビルド
    • IQ Bot

Image enhancement options in EnhanceImage action

Use the table in this topic to review all the available additional options to improve the quality of an image that you have uploaded in the EnhanceImage action.

Type Description
Grayscale Can be used to convert the colour pixels in an image to grayscale (Black & White).

Recommended: Convert colour images to greyscale to improve the OCR extraction accuracy.

Blur

Can be used to smoothen out areas of pixelation and improve image quality.

Recommended values: Can be any odd number (such as 3,11, 7), as long as it is lesser than the pixel size of the uploaded image.

Sharpen

Can be used to improve the clarity by removing any distorted or 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

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.
注: The principle use of the above 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

Can be used to smoothen out pixels that are not grayscale (black and white) to remove distortion or speckling from the image.

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

Contrast

Can be used to 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 Can be used to increase or decrease the brightness level of all the pixels in the image. Values can range from 0 (no change) to 100 (maximum contrast) or to -100 (minimum contrast).

Recommended values: From 0 through 100.

Thresholding Can be used to segment an image. For example, if the uploaded image has many elements, you can use this to locate specific objects or highlight boundaries within that image.

Recommended value: From 0 through 255.

Remove RGB Can be used to provide an upper and lower bound values to pixels in a colour image. Pixels in a colour image are represented by three primary colour channels: Red, Green and Blue (RGB). Output of this effect is a binary image having pixels that are Black (0) or White (255). Consider a scenario where you want to transform high intensity Red pixels represented by range (200,0,0) to (255,0,0) to White (255). You can set the lower bound for Red to (200,0,0) and upper bound to (255,0,0).
フィードバックを送る