Artificial neural networks based quantitative evaluation of subsurface anomalies in quadratic frequency modulated thermal wave imaging

Fuzzy Decision Tree Based Characterization of Subsurface Anomalies

“…Conventionally, pulse [2,[4][5] and lock-in stimulations [3] are famous stimulation mechanisms but these techniques limited by their high peak power stimulus and repeated experimentation respectively [3][4][5]. To overcome these limitations and to provide adequate defect depth resolution, Non-stationary thermal wave imaging systems have been introduced which uses moderate peak power optical stimulus modulated by a band of low frequencies [6][7][8]. Linear frequency modulation [6] and quadratic frequency modulated [7][8] stimulations are two varieties this NSTWI systems where QFMTWI probes more energy into the object at low frequencies compared to its linear counterpart.…”

“…To overcome these limitations and to provide adequate defect depth resolution, Non-stationary thermal wave imaging systems have been introduced which uses moderate peak power optical stimulus modulated by a band of low frequencies [6][7][8]. Linear frequency modulation [6] and quadratic frequency modulated [7][8] stimulations are two varieties this NSTWI systems where QFMTWI probes more energy into the object at low frequencies compared to its linear counterpart. On the acquired thermal response, variety of signal processing techniques employed for qualitative and quantitative characterization of the defects.…”

“…In Fuzzy decision tree, feature extraction and dimensionality reduction is carried out by random projection transform [13] and Fuzzification is performed by fuzzy c means clustering and to that FDT is employed on the fuzzy data set for defect classification and depth prediction. Further, a qualitative measure of detection is provided between afore mentioned and proposed methodology through defect SNR, sizing and probability of detection [8].…”

Numerical Simulation of Non-Destructive Characterization of CFRP Composite Through Frequency Modulated Thermal wave Imaging

“…Further, the application of compressed sensing [19] and sparse [20] based approaches improves the quality of thermograms and provide better understanding of defect detection. The emerging areas like machine Numerical Simulation of Non-Destructive Characterization of CFRP Composite Through Frequency Modulated Thermal wave Imaging learning [21,22] approaches further improves the detection and defect classification criteria's for NSTWI [23,24]. Further, in detail understanding of observed defects with their respective locations is presented by fusing the thermograms onto the original sample optical image [25].In present article, various post processing schemes have been analyzed for defect detection capability of FMTWI system for a CFRP with flat bottom holes of different sizes at different depths.…”

Bartlett Windowed Quadratic Frequency Modulated Thermal Wave Imaging

“…For in-detail visualization of defects, various image processing [19,20] techniques have been employed on obtained thermograms [21]. Further, machine learning [22][23][24] based processing modalities employed for defect detection in QFMTWI [25]. The present article mainly focused on the effects of spectral reshaping of thermal response from FMTWI system for a CFRP sample by employing Bartlett Window [17,18].…”