How to Use Image Filters in Cognex In-Sight Explorer
In-Sight Explorer has more than two dozen image filters to address the imaging challenges that occur in machine vision applications deployed in industrial automation environments. Multiple parameters enable application developers to fine-tune their filters to deal with issues like shadows, irregular lighting and unusual part shapes.
These are the filters mentioned in the attached training video:
Binarization: This is the most basic filter, providing a black-and-white image to work from. It’s the foundation of vision tools like Histogram and Blob. It compares each pixel in an image against a binarization threshold in the region of interest within the image.
Dilate. Using the Dilate filter enhances bright pixels in the image at the expense of dark pixels. This lightens an area of interest within an image. It’s useful for removing dark specks.
Erode. Using the Erode filter increases dark pixels and decreases light pixels. The opposite of Dilate, it darkens the region of interest.
Close. This filter combines the effects of the Dilate and Erode filters in sequential order, completely removing the dark areas of an image.
Open. This filter has the opposite effect of Close — first applying Erode and then the Dilate filter, which slightly decreases overall brightness in the image.
Fill Holes. The Fill Dark Holes and Fill Light Holes filters help close pixel gaps in the image.
High Pass. The High Pass filter emphasizes very fine features within the image. This helps accentuate edges.
Low Pass. The Low Pass filter has the opposite effect, smoothing the edges in the image and blurring them slightly.
Sharpen. This filter accentuates the edges of an image but produces a different effect than the High Pass filter.
Equalize. The Equalize filter extracts a histogram of the image and distributes the most common pixel intensity values. This gives equal pixel density across an image’s area of interest, which increases image contrast.
When running In-Sight Explorer, each filter is configured within a cell of your spreadsheet. It’s crucial to select the region of interest in the image you plan to enhance. This ensures the In-Sight Vision system applies filters only to the region of the image that requires correction or improvement, which helps to deliver the most effective enhancement and the fastest execution time in the production environment.
Note: These filters are only available on In-Sight Vision Systems running Firmware 5.1.0 or later. Find out more about firmware versions.