English title

---

Novel Ultrasonic imaging for weld inspection are based on employing ultrasonic phase arrays. The Total Focusing Method (TFM) is of recent imaging methods with utilizes the maximum available information achieved by array elements and cause a better defect detection compared with conventional methods. In contrast to its ability, TFM produces a considerable background noise which leads to a low resolution and accuracy of defect detection. In this paper an improved TFM method is presented that weights the received element signals using minimum variance theory.  In order to reduce the computational complexity and improve the detection accuracy, the weight computation is performed employing an iterative method such that in each iteration, the inverse of the correlation matrix resulted from the received signals is approximated in a recursive manner and then, is employed in the minimum variance based weight computation. Simulation results and experiments confirm the superiority of the proposed method.
Based on the simulation results, the proposed method reduces the average background noise about 16 dB compared to TFM and also improves the brightness of test block defects of ultrasonic image about 8 dB which means a better detection accuracy. Experiments verify the superiority of the proposed method, quantitatively and qualitatively. In addition, the runtime required for the proposed weighting operations is typically 10% of that for Gauss-Jordan removal technique applied to the weighted TFM imaging method.
 

---

Ultrasonic phased array imaging has found wide applications in industrial non-destructive test (NDT) including defect and crack detection in industrial samples and weld inspection. In this paper, an approximated minimum variance based beamforming method is proposed to improve the defect detection accuracy and reducing the background noise in the phased array ultrasonic imaging. The proposed method is employed in the boiler tube NDT. The matrix inverse approximation is computed using matrix inverse lemma such that the beamforming is performed proportional to the received signal amplitude. The performance of the proposed method is quantitatively and qualitatively evaluated in both theoretical and practical tests and compared with similar beamforming techniques. It is shown that the proposed method is able to reduce the image background noise while keeping the defect detection ability in comparison to both delay-and-sum and adaptive minimum variance beamforming methods. It also leads to defect detection enhancement compared with the minimum variance method. The computational complexity of the proposed method is less than that of both conventional and adaptive minimum variance beamforming methods.