Unsupervised learning-enabled pulsed infrared thermographic microscopy of subsurface defects in stainless steel.

Metallic structures produced with laser powder bed fusion (LPBF) additive manufacturing method (AM) frequently contain microscopic porosity defects, with typical approximate size distribution from one to 100 microns. Presence of such defects could lead to premature failure of the structure. In principle, structural integrity assessment of LPBF metals can be accomplished with nondestructive evaluation (NDE). Pulsed infrared thermography (PIT) is a non-contact, one-sided NDE method that allows for imaging of internal defects in arbitrary size and shape metallic structures using heat transfer. PIT imaging is performed using compact instrumentation consisting of a flash lamp for deposition of a heat pulse, and a fast frame infrared (IR) camera for measuring surface temperature transients. However, limitations of imaging resolution with PIT include blurring due to heat diffusion, sensitivity limit of the IR camera. We demonstrate enhancement of PIT imaging capability with unsupervised learning (UL), which enables PIT microscopy of subsurface defects in high strength corrosion resistant stainless steel 316 alloy. PIT images were processed with UL spatial-temporal separation-based clustering segmentation (STSCS) algorithm, refined by morphology image processing methods to enhance visibility of defects. The STSCS algorithm starts with wavelet decomposition to spatially de-noise thermograms, followed by UL principal component analysis (PCA), fine-tuning optimization, and neural learning-based independent component analysis (ICA) algorithms to temporally compress de-noised thermograms. The compressed thermograms were further processed with UL-based graph thresholding K-means clustering algorithm for defects segmentation. The STSCS algorithm also includes online learning feature for efficient re-training of the model with new data. For this study, metallic specimens with calibrated microscopic flat bottom hole defects, with diameters in the range from 203 to 76 µm, were produced using electro discharge machining (EDM) drilling. While the raw thermograms do not show any material defects, using STSCS algorithm to process PIT images reveals defects as small as 101 µm in diameter. To the best of our knowledge, this is the smallest reported size of a sub-surface defect in a metal imaged with PIT, which demonstrates the PIT capability of detecting defects in the size range relevant to quality control requirements of LPBF-printed high-strength metals.

Massive Ultrasonic Data Compression Using Wavelet Packet Transformation Optimized by Convolutional Autoencoders.

Ultrasonic signal acquisition platforms generate considerable amounts of data to be stored and processed, especially when multichannel scanning or beamforming is employed. Reducing the mass storage and allowing high-speed data transmissions necessitate the compression of ultrasonic data into a representation with fewer bits. High compression accuracy is crucial in many applications, such as ultrasonic medical imaging and nondestructive testing (NDT). In this study, we present learning models for massive ultrasonic data compression on the order of megabytes. A common and highly efficient compression method for ultrasonic data is signal decomposition and subband elimination using wavelet packet transformation (WPT). We designed an algorithm for finding the wavelet kernel that provides maximum energy compaction and the optimal subband decomposition tree structure for a given ultrasonic signal. Furthermore, the WPT convolutional autoencoder (WPTCAE) compression algorithm is proposed based on the WPT compression tree structure and the use of machine learning for estimating the optimal kernel. To further improve the compression accuracy, an autoencoder (AE) is incorporated into the WPTCAE model to build a hybrid model. The performance of the WPTCAE compression model is examined and benchmarked against other compression algorithms using ultrasonic radio frequency (RF) datasets acquired in NDT and medical imaging applications. The experimental results clearly show that the WPTCAE compression model provides improved compression ratios while maintaining high signal fidelity. The proposed learning models can achieve a compression accuracy of 98% by using only 6% of the original data.

Transmission of Images With Ultrasonic Elastic Shear Waves on a Metallic Pipe Using Amplitude Shift Keying Protocol.

Transmission of information using ultrasonic elastic waves on existing metallic pipes provides an alternative communication option across physical barriers in a highly partitioned industrial complex, such as a nuclear facility. This work investigates the feasibility of the transmission of digital images over metallic pipes. Ultrasonic communication systems for transmission of images on a nuclear-grade stainless steel pipe were assembled for bench-scale demonstration. Information carriers in this system are refracted shear waves transmitted and received with piezoelectric transducers (PZTs) operating at 2-MHz nominal frequency. The refraction and propagation of ultrasonic shear waves were modeled with COMSOL software. An amplitude shift keying (ASK) communication protocol for image transmission was developed and implemented in the GNURadio software-defined radio (SDR) environment. Digital information was converted to analog ultrasonic signals using Red Pitaya electronic boards. The performance of the ASK protocol is evaluated at the output of every block in the GNURadio program by monitoring the transmission of select characters. Using the ASK communication protocol, the transmission of the 32-KB image was demonstrated at 2-Kbps bitrate across 6-ft-long stainless steel pipe. Preliminary evaluation of ultrasonic communication on the piping of a nuclear facility, such as signal transmission on bent pipes, was performed with COMSOL computer simulations.

Introduction to the special issue on ultrasonics and ferroelectrics.

The sixteen articles in this special section were presented at the 2013 IEEE Ultrasonics, Ferroelectrics, and Frequency Control (UFFC) Symposium that was held in Prague, the Czech Republic, from July 21-25, 2013.

System-on-chip design for ultrasonic target detection using split-spectrum processing and neural networks.

Ultrasonic detection and characterization of targets concealed by scattering noise is remarkably challenging. In this study, a neural network (NN) coupled to split-spectrum processing (SSP) is examined for target echo visibility enhancement using experimental measurements with input signal-to-noise ratio around 0 dB. The SSP-NN target detection system is trainable and consequently is capable of improving the target-to-clutter ratio by an average of 40 dB. The proposed system is exceptionally robust and outperforms the conventional techniques such as minimum, median, average, geometric mean, and polarity threshold detectors. For realtime imaging applications, a field-programmable gate array (FPGA)-based hardware platform is designed for system-onchip (SoC) realization of the SSP-NN target detection system. This platform is a hardware/software co-design system using parallel and pipelined multiplications and additions for highspeed operation and high computational throughput.

Hardware-efficient realization of a real-time ultrasonic target detection system using IIR filters.

In this study, we address the increased computational demands of a frequency-diverse ultrasonic target detection system by developing a zero-phase IIR (ZP-IIR) filter. Several ZP-IIR filter types including Chebyshev-I, Chebyshev- II, and Butterworth were analyzed for their detection performance. The 4th-order filters with 8-bit quantized coefficients are shown to improve the flaw-to-clutter ratio by approximately 10 dB. Furthermore, the reduced adder graph algorithm is used for a hardware realization of ZP-IIR filters that does not require any dedicated multipliers. A small number of coefficients inherent to IIR filters and their multiplierless implementation provide efficient architecture suitable for compact, real-time ultrasonic imaging devices.

A successive parameter estimation algorithm for chirplet signal decomposition.

In ultrasonic imaging systems, the patterns of detected echoes correspond to the shape, size, and orientation of the reflectors and the physical properties of the propagation path. However, these echoes often are overlapped due to closely spaced reflectors and/or microstructure scattering. The decomposition of these echoes is a major and challenging problem. Therefore, signal modeling and parameter estimation of the nonstationary ultrasonic echoes is critical for image analysis, target detection, and object recognition. In this paper, a successive parameter estimation algorithm based on the chirplet transform is presented. The chirplet transform is used not only as a means for time-frequency representation, but also to estimate the echo parameters, including the amplitude, time-of-arrival, center frequency, bandwidth, phase, and chirp rate. Furthermore, noise performance analysis using the Cramer Rao lower bounds demonstrates that the parameter estimator based on the chirplet transform is a minimum variance and unbiased estimator for signal-to-noise ratio (SNR) as low as 2.5 dB. To demonstrate the superior time-frequency and parameter estimation performance of the chirplet decomposition, ultrasonic flaw echoes embedded in grain scattering, and multiple interfering chirplets emitted by a large, brown bat have been analyzed. It has been shown that the chirplet signal decomposition algorithm performs robustly, yields accurate echo estimation, and results in SNR enhancements. Numerical and analytical results show that the algorithm is efficient and successful in high-fidelity signal representation.

Ultrasonic data compression via parameter estimation.

Ultrasonic imaging in medical and industrial applications often requires a large amount of data collection. Consequently, it is desirable to use data compression techniques to reduce data and to facilitate the analysis and remote access of ultrasonic information. The precise data representation is paramount to the accurate analysis of the shape, size, and orientation of ultrasonic reflectors, as well as to the determination of the properties of the propagation path. In this study, a successive parameter estimation algorithm based on a modified version of the continuous wavelet transform (CWT) to compress and denoise ultrasonic signals is presented. It has been shown analytically that the CWT (i.e., time x frequency representation) yields an exact solution for the time-of-arrival and a biased solution for the center frequency. Consequently, a modified CWT (MCWT) based on the Gabor-Helstrom transform is introduced as a means to exactly estimate both time-of-arrival and center frequency of ultrasonic echoes. Furthermore, the MCWT also has been used to generate a phase x bandwidth representation of the ultrasonic echo. This representation allows the exact estimation of the phase and the bandwidth. The performance of this algorithm for data compression and signal analysis is studied using simulated and experimental ultrasonic signals. The successive parameter estimation algorithm achieves a data compression ratio of (1-5N/J), where J is the number of samples and N is the number of echoes in the signal. For a signal with 10 echoes and 2048 samples, a compression ratio of 96% is achieved with a signal-to-noise ratio (SNR) improvement above 20 dB. Furthermore, this algorithm performs robustly, yields accurate echo estimation, and results in SNR enhancements ranging from 10 to 60 dB for composite signals having SNR as low as -10 dB.