<abstract xml:lang="en" xmlns="http://www.w3.org/1999/xhtml"><p>In the current work, a filtering scheme for measurement of area and perimeter of heat affected zone (HAZ) in welding has been proposed along with identification of defects like porosity in the fusion zone. A filtering scheme based on wavelet filtering, edge detection and morphology has been designed and implemented on a welding image for this purpose. To see the effectiveness of the wavelet based proposed filtering scheme, SNR is calculated and compared with the filtered image processed without wavelet denoising. Experimental results revel that the SNR improves by 45-65% when wavelet filtering is introduced in the filtering scheme. Crisscrosses of the boundary of HAZ were observed and compared. The wavelet based proposed filtering scheme enhances the HAZ boundary smoothness by 12.5% in comparison to the filtering scheme without using wavelet denoising. In the limiting case, for poor quality image of weldment, the deviation in measurement of HAZ by manual and proposed scheme is within 10%. Error in manual evaluation significantly contributes in the above deviation as the HAZ on the base metal side in the raw image is diffused and difficult to measure accurately.</p></abstract>

In the current work, a filtering scheme for measurement of area and perimeter of heat affected zone (HAZ) in welding has been proposed along with identification of defects like porosity in the fusion zone. A filtering scheme based on wavelet filtering, edge detection and morphology has been designed and implemented on a welding image for this purpose. To see the effectiveness of the wavelet based proposed filtering scheme, SNR is calculated and compared with the filtered image processed without wavelet denoising. Experimental results revel that the SNR improves by 45-65% when wavelet filtering is introduced in the filtering scheme. Crisscrosses of the boundary of HAZ were observed and compared. The wavelet based proposed filtering scheme enhances the HAZ boundary smoothness by 12.5% in comparison to the filtering scheme without using wavelet denoising. In the limiting case, for poor quality image of weldment, the deviation in measurement of HAZ by manual and proposed scheme is within 10%. Error in manual evaluation significantly contributes in the above deviation as the HAZ on the base metal side in the raw image is diffused and difficult to measure accurately.

Implementation of Wavelet Denoising and Image Morphology on Welding Image for Estimating HAZ and Welding Defects Dibya

Manufacturing Management, BITS Pilani, Rajasthan, India

Precision Metrology Laboratory, Department of Mechanical Engineering, Sant Longowal Institute of Engineering and Technology, Longowal - 148106, India

Measurement Science Review

In the current work, a filtering scheme for measurement of area and perimeter of heat affected zone (HAZ) in welding has been proposed along with...

Implementation of Wavelet Denoising and Image Morphology on Welding Image for Estimating HAZ and Welding Defects Dibya

Data publikacji: 21 wrz 2011

Zakres stron: 108 - 111

DOI: https://doi.org/10.2478/v10048-011-0020-3

Słowa kluczowe
wavelet, edge detection, morphology, thresholding, heat affected zone (HAZ), porosity, signal to noise ratio

This content is open access.

Implementation of Wavelet Denoising and Image Morphology on Welding Image for Estimating HAZ and Welding Defects Dibya

Data publikacji: 21 wrz 2011

Zakres stron: 108 - 111

DOI: https://doi.org/10.2478/v10048-011-0020-3

Słowa kluczowewavelet, edge detection, morphology, thresholding, heat affected zone (HAZ), porosity, signal to noise ratio

This content is open access.

Słowa kluczowe
wavelet, edge detection, morphology, thresholding, heat affected zone (HAZ), porosity, signal to noise ratio