Automatic Detection and Segmentation of Laser Stripes for Industrial Measurement

Abstract

Laser stripes play an important role in industrial vision measurement as the major auxiliary feature. The existing research works mainly focused on the application of small size parts. However, with the increase of field of view (FOV), it is difficult to extract laser stripes robustly in varying field measurement situations because of the complex background, low proportion, and uneven characteristic of laser stripes. To increase the measurement adaptability in a complex environment, an automatic laser stripe detection and segmentation algorithm is proposed. First, the data set is constructed by a large number of image patches collected in the field and laboratory, and laser stripe patches in the imbalanced dataset are expanded by the data augmentation method. Next, the detection of the laser stripe is initially realized based on the training results of the convolutional neural network (CNN), and then the laser stripe is accurately detected by nonfeature filtering criteria based on area constraints. Finally, a subregional feature clustering method is proposed to realize effective segmentation of uneven laser stripes. A large number of verification experiments have been carried out in both the laboratory and field, and the results show that the proposed method can achieve automatic and accurate extraction of laser stripes, which has strong adaptability to both the complex background in the field and the uneven brightness characteristic of laser stripes, thereby satisfying the engineering requirements of large-scale part field measurement.