Learn More About X-ray Inspection


Explore the cutting-edge world of quality control with our latest video! Join us as we delve into the extensive process of using CT scans in the inspection of hypodermic needles. Discover how AI and computer vision algorithms transform raw data into actionable insights, automating quality decisions and ensuring top-notch production.

Don’t miss the impressive CT visuals revealing needle defects and the fascinating journey from scan to precision analysis. Subscribe now for a firsthand look at the future of X-ray inspection!

Remember to like & follow our socials for more content like this. Also, feel free to let us know what you think about CT inspections and even what kind of ordinary items you want to see a CT of.

CT can be a valuable piece of information if utilized effectively. Many companies use CT as the final output of their inspection process. However, it marks just the beginning of a comprehensive process involving computer fusion, artificial intelligence, and decision-making based on sample analysis.

Take, for instance, this sample we’ve picked up. Although we can’t reveal the actual customer’s product due to a non-disclosure agreement, it’s similar. Inside the box are 100 hypodermic needles. Traditional inspection would involve opening the box and examining each needle individually for bends or damage.

But we took a different approach. We performed a quick CT scan of the entire box, isolating each needle from the resulting volume. Subsequently, we applied computer vision algorithms to analyze each needle, calculating its specific parameters and determining whether it passed or failed quality control. We could even identify the location of a defective needle within the box for further action.

CT, in this context, is not the endpoint; it’s the starting point of the quality control process. From the CT volume, we run various AI algorithms to identify and evaluate each sample. This enables us to generate pass/fail decisions and automate actions like separating defective samples from the production line.

To illustrate, here are videos showing the computer tomography of the needle box. The needles rotate, and we can identify and categorize defects using colors like green, red, or blue. These defects include bends on the needle’s middle, base, or tip. All pass/fail determinations are made automatically by our quality control algorithm, leveraging computer vision.

By isolating each needle, we can cross-section it and measure the angle of any bends. A zero-degree angle is ideal, and anything over two degrees is considered a defect. This process streamlines the quality control journey, making it an essential starting point rather than a concluding one.

I hope you found this insight into the quality control process interesting! If you have more questions or want further details, feel free to ask. And don’t forget to subscribe for more content!