It has been several years since we started developing image comparison algorithms for a wide range of x-ray applications. Working on a Homeland Security program, the first application of these algorithms was the automated detection of counterfeit components, where we compared x-ray images of components in reels or trays looking for lot discrepancies. Since then, we have deployed image comparison algorithms for quality control in a variety of applications. A very popular use of image comparison algorithms, for example, is with our x-ray rental and custom units, where these algorithms serve to assist operators in finding defects. For this reason, an algorithm that can tell you what’s different from sample to sample is very powerful.
Although applications change, the premise is the same: a reference image is compared against a set of images that must look identical. Discrepancies must be carefully documented, as they likely signal defects.
The Image Comparison Algorithms for Radiography Unified Software – ICARUS – is the convergence of over a decade of algorithm development into a unified software platform. ICARUS combines several image comparison algorithms under one umbrella to give you the exact location in the x-ray images that differ from the reference image. ICARUS is a standalone software that works independently of other programs. However, when working in tandem with TruView 5 applications, it allows the user to stop automated acquisitions upon discovery of a fail.
ICARUS can find microscopic differences in the shape, size, and location of features, like broken or bent wire bonds.
To illustrate the power of ICARUS, the following example shows two images with small differences between them. The reference image on the left shows a module without any defects. The acquired image on the center presents a couple of subtle defects: a broken wire bond and a damaged solder ball. These defects were all correctly identified by ICARUS in the right image.
To improve the effectiveness of its algorithms, ICARUS allows you to select a comparison mask to focus analysis on regions of the image. For example, if you are interested in looking at the wire bonds only, all you need to do is to use the drawing tool in ICARUS to highlight the regions on the image that are of interest. ICARUS also has a powerful registration algorithm used to rotate and translate the acquired image to optimally match the reference image. ICARUS also outputs a complete report of all findings, including a PASS/FAIL result based on image similarity and a programmable threshold. ICARUS is currently in pre-release, and it is scheduled to start shipping November 1.
Please contact us today to see how ICARUS can improve the efficiency of your quality control program.