inspection method
Recently Published Documents

The wide application of intelligent manufacturing technologies imposes higher requirements for the quality inspection of industrial products; however, the existing industrial product quality inspection methods generally have a few shortcomings such as requiring many inspectors, too complicated methods, difficulty in realizing standardized monitoring, and the low inspection efficiency, etc. Targeting at these problems, this paper proposed an automatic detection and online quality inspection method for workpiece surface cracks based on the machine vision technology. At first, it proposed a vision-field environment calibration method, gave the specific method for workpiece shape feature recognition and size measurement based on machine vision, and achieved the on-line monitoring of workpiece quality problems such as feature defects and size deviations. Then, this study integrated the multi-scale attention module and the up-sampling module that can restore the locations of image pixels based on the high-level and low-level hybrid feature maps, built a workpiece crack extraction network, and realized workpiece crack feature extraction, crack type classification, and damage degree division. At last, experimental results verified the effectiveness of the proposed method, and this paper provided a reference for the application of machine vision technology in other fields.

Abstract 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.

Most of the bridge structures in the world are built of reinforced concrete. With the growth of service life and the increase of urban traffic and other factors, most bridges put into service have more or less damage. Traditional bridge damage detection methods include the manual inspection method and bridge inspection vehicle method, which have many shortcomings. Moreover, the detection of cracks in bridges is critical to the safety of transportation due to the extremely large number of bridges built in the road networks across the world. To this end, this paper uses the most widely used CNN in deep learning to identify and classify crack images and proposes a migration learning technique to solve the problem of the large amount of training data required for training CNN. The data augmentation and sliding window techniques are introduced to divide the collected crack data into training establish and test set. The experiments show that the method in this paper can classify the crack images better, extract and locate the cracks of bridge crack units, and finally extract the crack coordinates of boxing. Compared with the customary image recognition methods, the method used in this paper is easier to operate in practical engineering, and the accuracy of the obtained results is higher.

Rail corrugation appears as oscillatory wear on the rail surface caused by the interaction between the train wheels and the railway. Corrugation shortens railway service life and forces early rail replacement. Consequently, service can be suspended for days during rail replacement, adversely affecting an important means of transportation. We propose an inspection method for rail corrugation using computer vision through an algorithm based on feature descriptors to automatically distinguish corrugated from normal surfaces. We extract seven features and concatenate them to form a feature vector obtained from a railway image. The feature vector is then used to build support vector machine. Data were collected from seven different tracks as video streams acquired at 30 fps. The trained support vector machine was used to predict test frames of rails as being either corrugated or normal. The proposed method achieved a high performance, with 97.11% accuracy, 95.52% precision, and 97.97% recall. Experimental results show that our method is more effective in identifying corrugated images than reference state-of the art works.

Recently, there has been an increasing emphasis on the Indirect bridge health monitoring method employing passing vehicles, which is regarded as one of the most effective approaches in bridge damage screening. However,few researches have been conducted on the drive-by bridge inspection method using vehicle displacement profile as damage indicator. This paper proposes a new drive-by inspection method based on vertical vehicle displacementprofile with parameter optimization. A generalized Vehicle-Bridge Interaction (VBI) system is built in MATLAB, where the bridge is modelled as a simply supported beam with 10 elements, and the passing vehicle is represented as a simplified quarter car. To improve the result sensitivity to bridge damage, the parameter optimization of vehicle configuration is processed employing the Monte Carlo methods. Results show that the proposed method can successfully detect and localize bridge damage by using vertical vehicle displacement profile as damage indicator only, and its performance may depend on the vehicle configuration. The proposed approach provides merits in simplicity and efficiency, which can be applied widely to the bridge damage detection problems.

Abstract This paper proposes a method invention, namely an efficient NFT data inspection method with minimum granularity and probability comparison. The invention establishes a fast comparison method of AI model and data, that is, the direct comparison of small files priority and the maximum-minimum interval comparison. The invention takes the substantial identity inside the NFT data and the processing method of NFT data coincidence into account, so that the data content outside the token of the NFT publicly shared by the AI distributed system can also be unique on the Internet. Therefore, it can avoid the problem of incremental packaging and repeated packaging, and can successfully balance the efficiency and security of the comparison process. portions given in this document