Surface Defect Detection of Vehicle Light Guide Plates Based on an Improved RetinaNet

Aiming at the vehicle navigation light guide plate (LGP) image characteristics, such as complex and gradient texures, uneven brightness, and small defects, this paper proposes a visual inspection method for LGP defects based on an improved RetinaNet. First, we use ResNeXt50 with higher accuracy under the same parameters as the backbone network, and propose the lightweight module Ghost_module to replace the 1×1 convolution in the lower half of the ResNeXt_block. This can reduce the resource parameters and consumption, and speed up training and inference. Second, we propose and use an improved feature pyramid network (IFPN) module to improve the feature fusion network in RetinaNet. It can more effectively fuse the shallow semantic information and high-level semantic information in the backbone feature extraction network, and further improve the detection ability of small target defects. Finally, the defect detection dataset constructed based on the vehicle LGP images collected at a industrial site, and experiments are performed on the vehicle LGP dataset and Aluminum Profile Defect Identification dataset (Aluminum Profile DID). The experimental results show that the proposed method is both efficient and effective. It achieves a better average detection rate of 98.6% on the vehicle LGP dataset. The accuracy and real-time performance can meet the requirements of industrial detection.

A new extrusion die for aluminum profiles

A new type of extrusion die is put forward, in which cemented carbide material is used to inlay the working part of the die, so as to improve the wear resistance of the die. The necessity of using cemented carbide in extrusion die is introduced. The selection of cemented carbide was introduced by taking the actual round tube aluminum profile as an example. The method for determining the size of cemented carbide and the mosaic method of cemented carbide were described. Based on the results of extrusion, the common extrusion die and cemented carbide extrusion die were compared. The results show that the wear resistance of the die can be greatly improved by the use of cemented carbide. Thus the life of the die is greatly improved.

Research on Ultrasonic-Assisted Multi-Point Stretch Bending Process of Aluminum Profile

The tensile and bending process of asymmetric L-shaped aluminum alloy profile is studied by the Abaqus software using the finite element numerical simulation method. The geometric parameters of the ultrasonic-assisted vibration multi-point die (UMPD), and the law of influence on the stress-strain and spring-back of the L-section profile after bending are studied. The results show that the UMPD can reduce the forming stress of the profile during plastic deformation, and the stress-strain distribution of the aluminum profile is more uniform. The changes in the ultrasonic vibration frequency and amplitude of the mold are beneficial to reduce the spring-back of aluminum profiles. The ultrasonic process parameters with a vibration frequency of 20 kHz and an amplitude of 0.02 mm have the best effect on suppressing spring-back, which is reduced by 20.6% compared to the case of no ultrasonic application. Finally, it is verified by experiments that the experimental results are basically consistent with the simulation results, and the changing trend of spring-back deformation is consistent.

Effect of dual-convolutional neural network model fusion for Aluminum profile surface defects classification and recognition

<abstract> <p>Classifying and identifying surface defects is essential during the production and use of aluminum profiles. Recently, the dual-convolutional neural network(CNN) model fusion framework has shown promising performance for defects classification and recognition. Spurred by this trend, this paper proposes an improved dual-CNN model fusion framework to classify and identify defects in aluminum profiles. Compared with traditional dual-CNN model fusion frameworks, the proposed architecture involves an improved fusion layer, fusion strategy, and classifier block. Specifically, the suggested method extracts the feature map of the aluminum profile RGB image from the pre-trained VGG16 model's <italic>pool5</italic> layer and the feature map of the maximum pooling layer of the suggested A4 network, which is added after the Alexnet model. then, weighted bilinear interpolation unsamples the feature maps extracted from the maximum pooling layer of the A4 part. The network layer and upsampling schemes ensure equal feature map dimensions ensuring feature map merging utilizing an improved wavelet transform. Finally, global average pooling is employed in the classifier block instead of dense layers to reduce the model's parameters and avoid overfitting. The fused feature map is then input into the classifier block for classification. The experimental setup involves data augmentation and transfer learning to prevent overfitting due to the small-sized data sets exploited, while the K cross-validation method is employed to evaluate the model's performance during the training process. The experimental results demonstrate that the proposed dual-CNN model fusion framework attains a classification accuracy higher than current techniques, and specifically 4.3% higher than Alexnet, 2.5% for VGG16, 2.9% for Inception v3, 2.2% for VGG19, 3.6% for Resnet50, 3% for Resnet101, and 0.7% and 1.2% than the conventional dual-CNN fusion framework 1 and 2, respectively, proving the effectiveness of the proposed strategy.</p> </abstract>