No abstract

No abstract

The stress wave reflection method is widely used in the detection of structure size and integrity due to its advantages of low environmental impact and convenience. The detection accuracy depends on the accurate extraction of the stress wave reflection period. The traditional peak–peak method (PPM) measures the time interval between the first two peaks of the reflected waves to extract the reflection period. However, human interpretation is not avoidable for identifying the weak peak due to signal energy leaks into the surrounding environment. This paper proposes an algorithm for automatic extraction of the stress wave reflection period based on image processing to avoid human interference. The image is the short-time Fourier transform (STFT) spectrogram of the reflected wave signal after applying wavelet denoising and quadratic self-correlation operations. The edge detection method of image processing is used to extract the periodically occurring trough in the image. Graying and filtering are performed to eliminate interference. The frequency of the trough distribution is calculated by using the fast Fourier transform (FFT), and then the reflection period of the stress wave is obtained. The effectiveness and accuracy of the proposed method are validated by measuring the different lengths of two buried metal piles in soil. Comparing with the existing method of extracting the stress wave reflection period, this new algorithm comprehensively utilizes the time–frequency domain information of the stress wave reflection signal.

To improve the image quality of the deblurring results restored by existing blind deblurring method, an effective image blind deblurring method based on a single channel and L1-norm is proposed for the blurry images. Firstly, the blur influence on different channels in common colorspaces are analyzed and concluded to verify that the transformed channel for deblurring is reasonable. Then a more suitable fidelity term is proposed for increase image details and restraining ringing effects. Finally, the deblurring model in a single luminace channel is generated and solved by semi-quadratic regularization method. The comprehensive experiments on several representative blind deblurring methods to evaluate the performances in terms of objective evaluation, subjective vision and running time. The results have demonstrated that the propopsed method can ensure the deblurring performance and meanwhile decrease the running time. The deblurring performance of many different images reveals that our method is practical in image deblurring.

Printing defects are extremely common in the manufacturing industry. Although some studies have been conducted to detect printing defects, the stability and practicality of the printing defect detection has received relatively little attention. Currently, printing defect detection is susceptible to external environmental interference such as illuminance and noise, which leads to poor detection rates and poor practicality. This research develops a printing defect detection method based on scale-adaptive template matching and image alignment. Firstly, the research introduces a convolutional neural network (CNN) to adaptively extract deep feature vectors from templates and target images at a low-resolution version. Then, a feature map cross-correlation (FMCC) matching metric is proposed to measure the similarity of the feature map between the templates and target images, and the matching position is achieved by a proposed location refinement method. Finally, the matching image and the template are both sent to the image alignment module, so as to detect printing defects. The experimental results show that the accuracy of the proposed method reaches 93.62%, which can quickly and accurately find the location of the defect. Simultaneously, it is also proven that our method achieves state-of-the-art defect detection performance with strong real-time detection and anti-interference capabilities.