Machine learning and AI-created camera lenses capture recognized images ignoring certain objects. These less expensive optical sensors are built in hours using clear plastic and 3D printing.
Artificial Intelligence created a camera that includes 3D-printed plastic lenses and can capture specific object images, ignoring other frame objects. Without using power, selective lenses depend on complex plastic-printed patterns to diffract light in a unique method.
Professor Aydogan Ozcan from UCLA (University of California, Los Angeles), along with his colleagues, designed a deep-learning AI model that identifies number images of numbers from a huge online handwritten digit repository within a simulation.
The plasmonic sensing method of detecting small objects uses shining light on metal nanostructures to intensify the local electric field. Machine learning models feed plenty of data into an algorithm that learns to use its complex patterns and statistics to predict accurate outcomes better than traditional models. The UCLA team applied this concept to plasmonic sensing and created low-cost cameras with accurate sensor designs.
The lightweight and portable prototype include 3D printed plastic housing, four colorful light-emitting diodes (LED), and a camera. The AI algorithm uses LEDs and a computational method to measure the sensor output. Engineers and researchers can use this tool to enhance their optical sensor readers in healthcare and environmental monitoring applications.
These plasmonic sensors identify molecules of interest at a given electric field and can be “modified” to capture pictures of particular bio-targets like bacteria, viruses, cancer cells, etc.
An electrical engineering graduate student and Fellow of the National Science Foundation, Zach Ballard, first authored the study. He said, “Amazing discoveries and results are being made daily at research institutions like UCLA, but oftentimes when engineers begin to envision moving this science into the real world, they hit roadblocks. So it is always exciting to me to see cutting-edge technology become more practical.”