چکیده مقاله
Industrial radiography using gamma rays is one of the key tools for detecting internal defects in materials and welding joints However, image blurring caused by the inherent limitations of imaging systems can significantly hinder the precise analysis of defects, especially in welds In this study, a method based on the Sparse Land SL technique is proposed to mitigate blurring effects and enhance the quality of industrial radiographic images This method utilizes sparse signal reconstruction combined with human influenced image processing to reveal more details in critical welding areas In the first step, radiographic images are reconstructed using the Sparse Land algorithm and subsequently optimized through human centric processing techniques The Sparse Land method, relying on mathematical analysis and modeling of sparse signals, provides high accuracy in noise reduction and information recovery This capability is particularly valuable for images containing fine details and low contrast regions On the other hand, human influenced image processing, aimed at increasing the system's adaptability to industrial needs, improves defect detection accuracy in critical areas such as welding lines Experimental results indicate that this approach has improved image contrast by up to 30% To evaluate the method's performance, a set of radiographic images of welding joints under various conditions was analyzed The analysis of these images demonstrates that enhanced quality reduces detection errors and increases the reliability of radiographic analysis This is particularly significant in inspecting vital weld joints, where quality directly impacts structural safety The outcomes of this research could have substantial implications for sensitive industries such as oil and gas, aerospace, and construction By utilizing this method, not only is image quality improved, but inspection costs and time are also reduced These cost reductions are achieved through minimizing the need for repeated tests and speeding up analysis
کلیدواژهها
نویسندگان
Amir Movafeghi
Reactor and Nuclear Safety Research School, Nuclear Science and Technology Research Institute NSTRI , Tehran, Iran
Effat Yhaghi
Department of Physics, Faculty of Science, Imam Khomeini International University, Qazvin, Iran
Shabar Rasouli
Department of Physics, Faculty of Science, Imam Khomeini International University, Qazvin, Iran
شیوه ارجاع
Movafeghi, Amir and Yhaghi, Effat and Rasouli, Shabar,1403,Image Processing by the Sparse Land algorithm for Quality Improvement in Gamma-Based Industrial Radiography**,7th International Conference on Welding and Non Destructive Testing & 25th National Conference on Welding & Inspection & 14th National Conference on NDT & 3nd National Conference on Additive Manufacturing,Tehran
ارائهشده در
مجموعه مقالات هفتمین کنفرانس بین المللی جوشکاری وغیرمخرب، بیست و پنجمین کنفرانس ملی جوش و بازرسی، چهاردهمین کنفرانس ملی آزمایش های غیرمخرب و سومین کنفرانس ملی ساخت افزایشی1 اسفند 1403 · تهران