This paper focuses on compressive sensing (CS)-based hyperspectral imaging (HSI). A band-by-band reconstruction approach, namely prior image constrained compressive sensing (PICCS)-based HSI, is proposed. Furthermore, a more effective PICCS model is built in this paper. Each hyperspectral band is reconstructed based on the previous one, which utilizes not only the sparsity of each hyperspectral band in a certain basis but also the similarity between two consecutive bands. Moreover, compared with the algorithms which reconstruct all the hyperspectral bands simultaneously, PICCS-based HSI reduces the requirements for computational ability and computational memory of the receivers. In addition, compared with the independent band-by-band reconstruction algorithms and tensor-SL0-based HSI, PICCS-based HSI significantly reduces the number of measurements with similar or better reconstruction quality. The convergence of the two algorithms is proved and some simulations are provided to illustrate their effectiveness.
Time-frequency distribution (TFD) is two-dimensional function that indicates the time-varying frequency content of one-dimensional signals. And The Wigner-Ville distribution (WVD) is an important and effective time-frequency analysis method. The WVD can efficiently show the characteristic of a mono-component signal. However, a major drawback is the extra cross-terms when multi-component signals are analyzed by WVD. In order to eliminating the cross-terms, we decompose signals into single frequency components – Intrinsic Mode Function (IMF) – by using the Empirical Mode decomposition (EMD) first, then use WVD to analyze each single IMF. In this paper, we define this new time-frequency distribution as EMD-WVD. And the experiment results show that the proposed time-frequency method can solve the cross-terms problem effectively and improve the accuracy of WVD time-frequency analysis.
Human face localization in an image is a key component in many intelligent applications. A compact face detecting method, especially suitable for multiple faces in a complex background, is proposed in this paper based on the symmetry property of the face and facial organs. The detector has a low computation load, a simultaneous localizing capability and is robust to bad illumination. Experiments show the effectiveness of the algorithm so it is potentially useful in related intelligent applications.
Hybrid laser-arc welding is becoming one of the most significant laser welding technologies in industry due to its higher welding efficiency, higher tolerance to gaps between plates, and adjustment of composition and microstructure of the weld metal. Comparing with common off axis hybrid laser-arc welding, coaxially combined laser beam and arc can provide a symmetrical circular thermal source on the workpiece surface, which is convenient for 3-D welding. This paper introduces a coaxial hybrid CO2 laser-pulsed MIG welding system and conducts experiments of welding Al-Mg alloy plates under different welding conditions. The basic physical phenomena during welding are observed and the weld bead shape (penetration depth, weld width) are measured. The results show that hybrid laser-MIG can stabilize the arc, remarkably increase the total welding efficiency and improve the quality of weld bead formation. In addition, process and control techniques for hybrid laser-MIG welding are also proposed.
800MPa grade RPC steel is the recently developed new generation steel, which was produced by relaxation-precipitation controlling transformation (RPC) processing. The microstructure and mechanical properties of the heat affected zone (HAZ) in laser welding of 800MPa grade RPC steel were investigated by using a 4kW YAG laser and a Gleeble-1500 thermal simulator in this paper. The experimental results indicate that: (1) The size of prior austenite grains of HAZ increases with increasing heat input. (2) Granular bainite is the main microstructure of the HAZ of laser welding. (3) As the heat input increases, the hardness of HAZ decreases, but it is higher than that of the base metal, indicating no softened zone after laser welding. (4) As t8/5 increases, the impact toughness of HAZ increases at first and then decreases. The impact energy of HAZ is much higher than that of the base metal when t8/5 is between 3s and 8s. It indicates that excellent low temperature toughness can be obtained under appropriate laser welding conditions.
Evaluation of the cut quality is extremely significant to industrial applications of laser cutting. The relationship between cut quality and processing conditions has been investigated by using one of the measurable cut qualities, such as kerf width, striations, dross, roughness and so on. However, each of these qualities can only partially represent the cut quality. In this paper, a synthetic evaluation method for laser cutting quality has been proposed. A 3KW CO2laser was used to perform cutting experiments with 1.0mm thick mild steel sheets. The cut quality indicators, including kerf width, striations, dross, roughness, under different cutting conditions have been studied. A Synthetic Quality Number (SQN) has been presented as the evaluation indicator by quantitatively analyzing the conventional indicators. A neural network based method to anticipate laser cutting quality has been presented with SQN as the evaluation indicator.
This article designs a system for coaxially monitoring and penetration control in CO2 laser welding with a visual sensor. The key techniques for imaging the keyhole and eliminating the plasma-induced severe interference have been developed. The two-dimensional image of the keyhole has been obtained clearly in real time. It shows that once the penetration status changes from “partial penetration” or “weld pool penetration” (keyhole not penetrates workpiece) to “moderate full penetration” (keyhole penetrates workpiece), a lower grayscale spot appears in the center of the keyhole image and the grayscale curve presents a concavity in the center. This criterion of penetration recognition is suitable for most welding conditions. An image processing method that compares the average grayscale of a central image window at the keyhole center (vector Fc) and the average grayscale of an annular window surrounding the keyhole center (vector FR) has been proposed to quickly recognize “moderate full penetration”, based on which, the closed-loop control of penetration can be carried out and the weld bead can be kept at “moderate full penetration” during welding.
The theoretical and experimental study on the mechanism of plasma current during deep penetration laser welding is made in this paper. The plasma above the workpiece surface expands to nozzle, driven by the particle's consistency gradient, so an electric potential is formed between workpiece and nozzle due to the great difference in the diffusion velocities for ions and electrons. A kind of plasma-induced current can be obtained by circuiting the nozzle and workpiece, which current value is increased by adding a negative external voltage. As the external voltage increases, the plasma current increases until to a certain saturation value. As the workpiece-to-nozzle distance increases, the plasma current decreases. The decreasing rate becomes lower gradually until the plasma cunent becomes zero. Based on above analyzing, closed-loop control of workpiece-to-distance was realized based on the monotonous relationship between the plasma current and the distance. By applying a transversal magnetic field, the plasma above the keyhole can be somewhat eliminated, therefore the attenuation of plasma to the incident laser power can be lessen and the penetration depth can be increased.
In underwater laser beam welding, the shielding condition of the local dry cavity is one of the key factors affecting the weld quality. The variation in the weld bead formation and mechanical properties of the welded joint under various shielding conditions was investigated in this paper. The welding experiments were performed with Type3O4 stainless steel as base metal and Type-ULC308 electrode as the filler wire by using a 4kW Nd:YAG laser. The experimental results show that variation of gas flow speed has important influence on the weld bead formation. The bead width and reinforcement of the welds vary slightly under different gas flow speed, but the penetration depth is much lower under higher gas flow speed. Lower gas flow speed induces poor shielding conditions and increases the oxygen content of weld metal up to 800ppm, but higher gas flow speed that may form an excellent local dry cavity could prevent the weld metal being oxidized, with only about 8Oppm in oxygen content. Although the tensile strength of underwater laser welds keeps constant no matter how the shielding conditions are, the ductility of weld metal decreases largely for the weld obtained in poor shielding conditions comparing with that of good shielding conditions. The structure of the shielding nozzle has more important effect on the shielding condition, in specific, appropriately enlarging the outside diameter can considerably improve the stability of the local dry cavity.
To guarantee good quality of butt laser welding with filler wire, a sensor, that is used to detect gap width and seam tracking and a wire feeding system have been developed in this paper. The sensor contains a laser diode as assistant optical source and a photoelectric cell as receiving component. A step-motor drives the sensor transversally scanning above the work piece so as to detect the gap width as basis of adjusting welding parameters and the center position of the butt gap as benchmark of seam tracking. A computer calculates the required wire feeding speed and welding speed according to the gap width and transfers these values to wire feeding system and speed adjusting system. In the welding of 2mm-thick low carbon steel sheet, the wire feeding speed and welding speed are respectively changed according to gap width detection and weld bead appearance is analyzed. The results show that it is better to adjust wire feeding speed for narrower gap and adjust welding speed for wider gap. According to the detected signal of the sensor, the appearance quality of the total weld could be guaranteed by choosing wire feeding speed or welding speed as the control variable for different gap widths.
KEYWORDS: Laser welding, Signal processing, Plasma, Picosecond phenomena, Control systems, Signal detection, Process control, Sensing systems, Sensors, Laser processing
A sensing and control system of process quality in CO2 laser deep penetration welding has been developed, including plasma photo sensor (PS) and plasma charge sensor (PCS), signal processing circuit, personal computer, and optimization and control software. By using the information of plasma radiation and plasma charge, the focusing lens is adjusted to the optimum focal point position (at which the maximum penetration can be achieved) and closed-loop control is realized to assure the focal point position and penetration depth constant. In laser welding of the workpiece aslant placed with larger angle, the variation of focal point position is less than 0.2 mm and the fluctuation of penetration depth is less than 0.05 mm. The control system has been preliminarily applied in laser precision welding of 2.5 m long zircaloy fuel channel used in nuclear heat-supply reactor.
In high-power CO2 laser welding, in addition to well- known stable heat conduction welding (HCW) and stable deep penetration welding (DPW), the authors have found a third welding process: unstable-mode welding (UWM). UMW has the feature of the two welding modes (HCW and DPW) appearing intermittently. The physical characteristics of the three welding processes and the features of their weld formation have been described. By synthesizing the effects of laser power, focal position and travel speed on welding mode and welding process, the laser welding mode-transition curves which include three kinds of welding process have been obtained. The curves precisely indicate the welding- parameter ranges of three different welding processes. During laser welding, the thermal focusing of the lens will change the focal position and may cause transition of welding mode and fluctuation of penetration and weld width. The relationships between lens focal shift, laser power and beam irradiation time have been investigated. By using welding mode-transition curve, the influence of thermal focusing on welding mode and welding process has been studied, and measures for prevention of welding process instability induced by thermal focusing have also been suggested.
An advanced plasma detecting system of CO2 laser welding has been developed. The system consists of three sensors, signal processing, A/D data converting and photo-electric coupling units connected with a rapid personal computer. The Photocell Sensor (PS) detects the intensity of the blue light irradiated by the plasma. The Plasma Charge Sensor (PCS) detects the electric density of the plasma plume. The Microphone Sensor detects the sound pressure coming from the rapidly expanding vapor in the keyhole. All of the sensors can exactly distinguish three kinds of welding processes--heat conduction welding, deep penetration welding, unstable mode welding. When the welding parameters are given, the PCS signal depend on the distance between the welding nozzle and the workpiece, the PS signals are correlated closely to the focal point position. Three sensors can be used to control the focal point position (penetration depth) under given laser power and welding speed. In addition, the relation between detecting signals and penetration depth is given. The sensors of the system have features of simple structure, low cost and high sensitivity, which are especially suitable for on-line plasma detecting, quality controlling and off-line plasma analyzing of CO2 laser welding.
The authors have found a third process--unstable-mode welding (UMW) under a certain condition, besides already known stable deep penetration welding (DPW) and stable heat conduction welding (HCW) during high-power CO2 laser welding. UMW has basic feature that the two welding modes (DPW and HCW) appear intermittently, with the penetration depth and weld width jumping between large and small grades. In this paper, the physical phenomena, especially the signal of plasma during welding and the weld-forming of three kinds of process have been investigated. Effects of welding parameters (focal position, laser power and travel speed) on laser welding mode and weld-forming have been comprehensively studied. Double-U curves of laser welding mode transition have been obtained, which indicate the parameter ranges of the three monitored welding process.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.