Industrial Inspection with Open Eyes: Advance with Machine Vision Technology

Liu, Zheng; Ukida, Hiroyuki; Niel, Kurt S.; Ramuhalli, Pradeep

doi:10.1007/978-1-4471-6741-9_1

Cited by 6 publications

(3 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nowadays, thermal imaging in the midwave infrared mainly relies on bulk semiconductors with a narrow bandgap . Mercury cadmium telluride (MCT), indium antimonide (InSb), and superlattice structures are all mature materials for focal plane array imaging chips with good quantum efficiency; however, the high fabrication and processing costs hinder potential applications such as autonomous vehicles, face recognition, and industrial inspection . Cheaper materials with fast response, high sensitivity, and compatibility with large-scale production are the key to facilitate civilian applications.…”

mentioning

confidence: 99%

Thermal Imaging with Plasmon Resonance Enhanced HgTe Colloidal Quantum Dot Photovoltaic Devices

2018

View full text Add to dashboard Cite

Thermal imaging in the midwave infrared plays an important role for numerous applications. The key functionality is imaging devices in the atmospheric window between 3 and 5 μm, where disturbance from fog, dust, and other atmospheric influence could be avoided. Here, we demonstrate sensitive thermal imaging with HgTe colloidal quantum dot (CQD) photovoltaic detectors by integrating the HgTe CQDs with plasmonic structures. The responsivity at 5 μm is enhanced 2- to 3-fold over a wide range of operating temperatures from 295 to 85 K. A detectivity of 4 × 10 Jones is achieved at cryogenic temperature. The noise equivalent temperature difference is 14 mK at an acquisition rate of 1 kHz for a 200 μm pixel. Thermal images are captured with a single-pixel scanning imaging system.

show abstract

mentioning

confidence: 99%

Thermal Imaging with Plasmon Resonance Enhanced HgTe Colloidal Quantum Dot Photovoltaic Devices

2018

View full text Add to dashboard Cite

show abstract

“…There is precedent for visual condition assessment applications, with emerging low-cost, low-power mobile processors making it possible to implement computer vision on portable devices [7].…”

Section: Prior Artmentioning

confidence: 99%

“…Further, as deep learning becomes prevalent, it is finding use in visual condition monitoring, where it is applied to capture insights from difficult-to-comprehend transformations, even transitioning from diagnostics to prognostics. Such advances have been leveraged in conjunction with infrared imaging to detect wear and other issues within mechanical systems [8,9], to inspect vehicle component production [7], and to automate monitoring of rail surface defects where data volume prohibits manual inspection [10].…”

Section: Prior Artmentioning

confidence: 99%

Automotive Diagnostics as a Service: An Artificially Intelligent Mobile Application for Tire Condition Assessment

Siegel

Sun

2018

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Automotive tires must be maintained to ensure vehicle performance, efficiency, and safety. Though vehicle owners know to monitor tread depth and air pressure, most are unaware that degrading rubber poses a safety risk. This paper explores the need for tire surface condition monitoring and considers the development of a densely connected convolutional neural network to identify cracking based on smartphone images. This model attains an accuracy of 78.5% on cropped outsample images, besting inexperienced humans' 55% performance. Using this model, we develop a web service allowing mobile devices to upload images captured by their internal cameras for online classification, as the basis of an AI-backed "Diagnostics-as-a-Service" platform for vehicle condition assessment. By encoding knowledge of visual risk indicators into a neural network model operable from a user's trusted smartphone, vehicle safety may be improved without requiring specialized operator training.

show abstract