Artificial intelligence, or AI, has seen a recent surge in use across a wide variety of industries. Anyone who has used Google or Amazon has been subject to one of the more popular uses of AI in the customized advertising. Another popular use of AI is image processing, which has seen an increased use in intelligence gathering and other security applications. Deep learning, a subset of AI, is particularly useful in image processing.
Deep learning algorithms use neural networking to process enormous data sets in order to carry out a desired task. For example, Japanese researchers used images of healthy tissue and cancerous tissue to teach a computer to detect early stage cancer using a deep learning algorithm. The computer detected cancerous tissue successfully about 80 percent of the time, which is comparable to a standard medical diagnosis.
Even beyond its security and medical applications, it is easy to see how deep learning could be applied to industrial vision systems. In very simple terms, a deep learning algorithm could automate quality control by learning what a “good” part looks like, and rejecting the rest. More broadly, deep learning algorithms could be used to learn what part of the manufacturing process causes a set of defects. The algorithm could then adjust the process and remove the defect from subsequent parts on the fly.
A simple example of process improvement is cookie baking. The deep learning algorithm could learn what a perfect cookie is, what a burned cookie is, and what an under-cooked cookie is. By actively monitoring the cookie batches using a vision system, the algorithm would be able to adjust the oven temperature as the cookies were moving through the process, saving time and money on bad batches.
The applications of deep learning are almost limitless with machine vision system. The ability of AI to process images is a well documented use of the technology, and most manufacturers have large databases of past material that could easily be used by deep learning algorithms for initial learning. Once the algorithm has learned, it can be maintained and adjusted over time just like any other piece of industrial equipment.
If you are interested in learning more about deep learning and its application to machine vision systems, contact us for a consultation today.
Learn more about Deep Learning Applications and Functions below.
Machine Vision and Deep Learning A Machine Vision Revolution. Machine vision has become widely accepted by multiple industries as the most efficient method of detecting flaws, errors, and irregularities in manufactured goods. Certain types of inspection have been, until recently, impossible for automated systems due to the complexity and nuance required to distinguish between cosmetic […]
Increase Food Inspection and Production Levels with Advanced Machine Vision AI’s higher intelligence is like having Superman join your quality inspection team. Imagine having the capability to scan 1,000 parts per minute! Machine vision equipped with deep learning can accomplish this task. That breaks down to inspecting one part in approximately six milliseconds. With machine […]
Time and Money Saved from Machine Vision and Deep Learning Critical Inspections Beyond Human Capabilities We all know how crucial machine vision inspections are to every type of manufacturing business in our worldwide market. Competition is too high to allow defective parts and products to hit the shelves to consumers. Not to mention the legality […]
Artificial intelligence (or AI) has seen a recent surge in use across a wide variety of industries. One type of AI currently being used in image processing is called Deep Learning. Deep learning algorithms use neural networking to process enormous data sets in order to carry out a desired task. The applications of deep learning […]
Deep learning is currently being used in the automotive industry for a number of inspection applications. Most manufacturers have large databases of past material that can easily be used by deep learning algorithms for initial learning. Once the algorithm has been learned, it can be maintained and adjusted over time, just like any other piece […]
Machine Vision and Deep Learning Deep learning is becoming a commonly heard word in business, science and programming circles. However, the concept is still quite new to many people. This important type of algorithm is a subset of machine learning. It uses neural networks and pattern recognition to develop deep insights that would ordinarily be […]
Machine Vision and Deep Learning Artificial intelligence, or AI, has seen a recent surge in use across a wide variety of industries. Anyone who has used Google or Amazon has been subject to one of the more popular uses of AI in the customized advertising. Another popular use of AI is image processing, which has […]
CHALLENGES IN PAD PRINTING INSPECTION Multi-step tampography can lead to considerable but tolerable relative shifts (registration problem) Varying quantities of ink applied results in fonts or lines that are thicker/thinner in appearance Random texture of substrate, like brushed or otherwise decorated metal With Cognex ViDi Suite, the automated aesthetic inspection of complex pad printing […]
CHALLENGES IN MACHINED PART INSPECTION Typically there are many different types of complex shapes Varying surface properties depending on machining tool quality and varying properties of the blank material need to be tolerated Some defects only show under a very particular combination of illumination, camera and surface orientation With ViDi Suite, the automated visual inspection […]
Update: ViDi is now owned and operated by Cognex Corporation Integro Technologies, a premier machine vision integrator, announces its partnership with ViDi Systems SA, an innovative software firm in Switzerland. ViDi is revolutionizing the industry by commercializing the first Deep Learning-based image analysis solution dedicated to the machine vision market. ViDi recently designed and developed […]
CHALLENGES IN AUTOMATIC DETECTION Typical defects like hits, scratches or stains are often hardly discernible at the early production stages because of the parts’ rough and strongly textured surfaces. Normal variations and non-significant anomalies in the material need to be tolerated. Stains & hits often manifest themselves by a local change in contrast caused by […]