- Fourier/Frequency: Ultimate Sharpener • Pattern Remover • Interactive Deblur • Freq. Filters
- Image Adjustment: Color Safe Levels and Curves • Adaptive Equalization • Local Equalization
- Noise Removal: Frame Averaging • Universal Noise Remover (8 different filters) • JPEG Cleaner
- Sharpening: Detail Sharpener • Edge Enhancer • Double Unsharp Mask
- Advanced/Utility: Checksum • Image Math • Custom Matrix Convolution
Ultimate Sharpener – Ultimate Sharpener will normalize the brightness in an image and increases the contrast at the same time.
Pattern Remover – Pattern Remover is ideal for latent print enhancement. Simpy select the background pattern you want removed and pattern remover does the rest. Pattern remover is an automated process with interactive sliders and real time preview so you can see the effects of your adjustments as you make them. This automated pattern remover also works in color.
Interactive Deblur – Fourier image processing works in the frequency domain rather then the pixel domain. This allows us to do image processing in a much more efficient manner. Previous attempts at working in the frequency domain required the user to have a strong math background to properly understand the process. With ClearID Interactive Deblur, those concepts have been integrated into a few slider bars that allow the user to make adjustments and see the effects of changing the settings. Interactive Deblur also Includes Focus Deblur and Motion Deblur.
Frequency Filters – The Frequency Filters plug-in builds on the Pattern Remover and offers the user a way to learn more about how the Power Spectrum is laid out and the effects of high pass/low pass filtering. We can actually generate our own masks in the frequency domain using three different types of masks.
Image Adjustment Tools
Color Safe Levels – Color-Safe Levels enhances under/overexposed images. There is a histogram adjustment on the bottom left hand side of the dialog. There is also a drop down towards the bottom right, where the user can use the zone system (Developed by Ansel Adams) to adjust gamma in the image.
Color-Safe Curves – Color-Safe Curves allows the user to adjust contrast levels without affecting the color saturation of the image. This plug-in is intended to replace the Curves plug-in from Photoshop. It does not use the LAB color space to adjust the histogram and saturation of the image. The LAB color space does not handle color saturation properly. When the user starts to brighten the image it will desaturate the image color. When the user starts to darken the image it saturates the image color.
Adaptive Equalization – Adaptive Equalization is used to help to make fine details visible in an image while somewhat suppressing the effects of noise in the image. This can be useful when there are wide ranges of brightness and darkness in an image.
Local Equalization – Local Equalization is used to help to make details visible in an image but also brings in the noise. This takes pixel values that are slightly brighter (or darker) than their surrounding pixel values and makes them slightly brighter (or darker) to increase their contrast.
JPEG Cleaner – JPEG Cleaner reduces the block artifacts that JPEG compression can cause. Many processing techniques can exaggerate the blockiness of the JPEG compression. Applying JPEG Cleaner first can reduce the artifacts caused in those processing techniques. In some cases if the compression quality is very low, the user may want to apply JPEG Cleaner twice.
Frame Averaging – The Frame Averaging plug-in allows the user to perform frame averaging over many frames, currently up to 32,767 images. This will allow the user to remove noise that appears over several images.
Universal Noise Removal – includes the following:
- Average Blur, especially in the Round case, is very good for dealing with empty magnification – images that have been blown up too far.
- Gaussian Blur, This is a softer blur then the Average blur. It is not good for preserving edges but will reduce noise effectively. This is the same as the Photoshop Gaussian Blur function and is here for comparison purposes. The Photoshop version operates on the individual RGB channels. The Neighborhood Radius determines the size of the blur applied to the image. The larger this value the more detail will be blurred and the more noise that will be removed
- Non-Edge Blur, blurs the non-edge areas. This will create an artificially smooth image, but with an opacity change, you can back off the non-edge detail to help bring out the non-edge areas.
- Conditional Smooth is good for preserving edges while still reducing noise.
- Classical Median can use a round neighborhood instead of only a square neighborhood, as in Photoshop’s Median. This produces a nicer looking image and preserves the color of the image correctly.
- Hybrid Median is an edge-preserving and corner-preserving median filter.
- Vertical Filter 3 & 5 are very useful for removing certain kinds of interlace artifacts.
- The Effect Opacity (%) will act as a blend function to compare it to the original image.
Detail Sharpener – Detail Sharpener brings out fine details without over-sharpening edges.
Edge Enhancer – Edge Enhancer will sharpen edges without increasing noise. Remember, when sharpening an image if noise is not reduced first, it will be enhanced along with the edges. Edge Enhancer uses a technique called Difference of Gaussians. This is similar to unsharp masking. The image is processed with two different Gaussian kernels. The images are subtracted from each other and the difference between those two kernels is the result.
Double Unsharp Mark
The Double USM (or Double UnSharp Mask) filter allows separate control of the shadows and highlights. Two things happen in a traditional UnSharp Mask:
1) The transition between dark and light gets steeper (increasingly steeper as the radius increases)
2) The overshoot gets bigger – really two parts – how tall it is and how wide it is. That overshoot is really the halo artifact that we’re familiar with.
Checksum – The Checksum plug-in generates a SHA1 value for the currently selected image and layer. SHA (Secure Hash Algorithm) is a cryptographic hash function. This allows us to create a 160 bit alphanumeric sequence that uniquely identifies an image and or layer in that image. This will be very useful to verify that a plug-in with the same group of settings generates repeatable results. This plug-in can also be used for verification purposes. A user can submit their settings and SHA1 value to another user to see if they can verify their results on an image.
The script for this plug-in works in an automated fashion. The user can apply this script to their image and it will create a .log file with information about their image. The .log file will be located in the same directory as the image and will be the same name as the image with a .log file extension. For example, if the image is called ocean.tif, the log file would be found under the same directory and be called ocean.tif.log. The log file will open up in most text editors such as Word or Notepad. The log file will contain information about each layer in the image.
Image Math – Image Math can be used to perform various mathematical operations on an image. Most likely this plug-in will be combined with scripts or actions to automate certain functions.
- Store (Image Memory) – This stores the current image into a memory buffer to be used in a later operation. For example if a user wishes to add one image to another, they can store an image into the memory buffer. Bring up another image, open the Calculations plug-in again and this time choose Add to add the two images together.
- Clean (Image Memory) – This clears the current image in the memory buffer.
- Add – This selection can be used with the Store function to add two images together. This will in effect brighten an image.
- Subtract – This selection can be used with the Store function to subtract the stored image from the current image. This can be useful to see what is different from one image to another or for background subtraction.
- Multiply – This multiplies two images together. This will have a similar effect to the Add selection, although a bit more severe, and brighten the two images.
- Divide – This divides two images together. This will divide the current image from the current image. This will be similar to the Subtract selection and will show items that have changed.
- Absolute Diff. – This is similar to the Subtract option. This will subtract one image from the other, if the result is a negative number, e.g. 40-150 = -110, the Absolute Difference of that would be 110.
- Darker – This compares two images and takes the darker of the two pixel values at each pixel.
- Lighter – This compares two images and takes the lighter of the two pixel values at each pixel.
- Primary Image Only – This refers to the current image in the preview window. A Divide By or Add Constant can be added to the current image only.
- Divide By – This is a scale correction that is applied to the result.
- Add Constant – This is an offset that can be applied to the image.
Custom Matrix Convolution – The Custom Matrix Convolution filter allows the user to perform a convolution technique using user-defined kernels. Each pixel is replaced with an average of itself and its neighbors. We can define up to a 7×7 array of values to perform the kernel operation. As the kernel becomes larger the user may wish to use a non-square or circular weighting. The end result can be a noise reduction in the image, edge detection or even sharpening of an image depending on the type of kernel used. This is similar to the Photoshop Custom filter but has a larger 7×7 kernel size compared to the 5×5 kernel in Photoshop.
« Back to the Knowledgebase