Application of attention-DnCNN for ESPI fringe patterns denoising
Linlin Wang, Run Li, Feng Tian, Xiaoyu Fang
Abstract
Fringe patterns' denoising in electronic speckle pattern interferometry (ESPI) is an important step in phase extraction. In this study, we propose a new denoising method for ESPI fringe patterns based on a convolutional neural network (CNN). The proposed model which combines the attention mechanism and CNN is defined as attention-denoising CNN. In this model, owing to the attention mechanism, more attention will be paid to fringe information, and better filtering results will be achieved. The experimental results show that our proposed method can obtain excellent results, especially with high and large variation density ESPI fringe patterns.
Topics & Concepts
Computer scienceElectronic speckle pattern interferometryConvolutional neural networkNoise reductionSpeckle patternArtificial intelligenceSpeckle noisePattern recognition (psychology)Computer visionOptical measurement and interference techniquesImage Processing Techniques and ApplicationsAdvanced Measurement and Detection Methods