feature recognition
Recently Published Documents





2022 ◽  
Vol 73 ◽  
pp. 102260
Peizhi Shi ◽  
Qunfen Qi ◽  
Yuchu Qin ◽  
Paul J. Scott ◽  
Xiangqian Jiang

2022 ◽  
Vol 2022 ◽  
pp. 1-11
Rui Yang ◽  
Zenghui An ◽  
Shijun Song

A convolutional neural network has the characteristics of sharing information between layers, which can realize high-dimensional data processing. In general, the convolutional neural network uses a feedback mechanism to realize parameter self-regulation, which solves the disadvantages of manual parameter adjustment. However, it is unable to determine the iteration number with the best calculation accuracy. Calculation efficiency cannot be guaranteed while achieving the best accuracy. In this paper, a multilayer extreme learning convolutional neural network model is proposed for feature recognition and classification. Firstly, two-dimensional spatial characteristics of planetary bearing status data were enhanced. Then, extreme learning machine is embedded in a convolution layer to solve convex optimization problems. Finally, the parameters obtained from the training model were nested into a network to initialize the model parameters to separate each status feature. Planetary bearing experimental cases show the effectiveness and superiority of the proposed model in the recognition and classification of weak signals.

Cuili Mao ◽  
Wen Ma

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.

2022 ◽  
Vol 2022 ◽  
pp. 1-9
Li Xue ◽  
Chuangjian Yang

In order to improve the effect of copying and recreation of painting works, this paper combines mobile digital multimedia big data technology to improve the image coding algorithm, identify the characteristics of existing works, apply the algorithm to the detailed analysis of painting works, and construct the main functional structure modules of the system. Moreover, this paper combines the existing hardware equipment to construct the painting works’ recreation system and obtains the image processing module. After the system is constructed, the effect of copying and recreating painting works is analyzed through the mobile digital multimedia big data analysis technology. Finally, this paper constructs the system of this paper through simulation methods and uses experiments to calculate the feature recognition effect and copy effect of the painting works of the system. Through experimental analysis, it can be known that the copying and recreation system of painting works based on mobile digital multimedia big data analysis proposed in this paper can help painters effectively improve the effect of recreation.

2022 ◽  
Vol 2146 (1) ◽  
pp. 012037
Ying Zou

Abstract Aiming at the problems of high complexity and low accuracy of visual depth map feature recognition, a graph recognition algorithm based on principal component direction depth gradient histogram (pca-hodg) is designed in this study. In order to obtain high-quality depth map, it is necessary to calculate the parallax of the visual image. At the same time, in order to obtain the quantized regional shape histogram, it is necessary to carry out edge detection and gradient calculation on the depth map, then reduce the dimension of the depth map combined with the principal component, and use the sliding window detection method to reduce the dimension again to realize the feature extraction of the depth map. The results show that compared with other algorithms, the pca-hodg algorithm designed in this study improves the average classification accuracy and significantly reduces the average running time. This shows that the algorithm can reduce the running time by reducing the dimension, extract the depth map features more accurately, and has good robustness.

Materials ◽  
2021 ◽  
Vol 15 (1) ◽  
pp. 247
Xinyi Xiao ◽  
Hanbin Xiao

Robotic additive manufacturing (AM) has gained much attention for its continuous material deposition capability with continuously changeable building orientations, reducing support structure volume and post-processing complexity. However, the current robotic additive process heavily relies on manual geometric reasoning that identifies additive features, related building orientations, tool approach direction, trajectory generation, and sequencing all features in a non-collision manner. In addition, multi-directional material accumulation cannot ensure the nozzle always stays above the building geometry. Thus, the collision between these two becomes a significant issue that needs to be solved. Hence, the common use of a robotic additive is hindered by the lack of fully autonomous tools based on the abovementioned issues. We present a systematic approach to the robotic AM process that can automate the abovementioned planning procedures in the aspect of collision-free. Typically, input models to robotic AM have diverse information contents and data formats, hindering the feature recognition, extraction, and relations to the robotic motion. Our proposed method integrates the collision-avoidance condition to the model decomposition step. Therefore, the decomposed volumes can be associated with additional constraints, such as accessibility, connectivity, and trajectory planning. This generates an entire workspace for the robotic additive building platform, rotatability, and additive features to determine the entire sequence and avoid potential collisions. This approach classifies the uniqueness of autonomous manufacturing on the robotic AM system to build large and complex metal components that are non-achievable through traditional one-directional AM in a computationally effective manner. This approach also paves the path in constructing an in situ monitoring and closed-loop control on robotic AM to control and enhance the build quality of the robotic metal AM process.

2021 ◽  
Vol 2021 ◽  
pp. 1-9
Weiwei Luo

In order to deal with the problem that the traditional stage costume artistry analysis method cannot correct the results of big data clustering, which leads to deviations in the extraction of costume artistry features, this paper proposes a clothing artistic modeling method based on big data clustering algorithm. The proposed method provides a database for big data clustering by constructing the attribute set of the big data feature sequence training set and, at the same time, constructing a second-order cone programming model to correct the big data. Aiming at the problem that traditional stage costume art analysis methods cannot correct the clustering results of big data. On this basis, the costume elements of the opera stage are segmented, initialized, and transformed into a binary function. Finally, using the convolutional neural network, combining the element segmentation results and the large data clustering space state vector, a feature extraction model of stage costume art is constructed. Experimental results show that the model has good convergence, short time-consuming, high accuracy, and ideal feature recognition capabilities.

2021 ◽  
Vol 12 (1) ◽  
pp. 3
Abdullah D. Ibrahim ◽  
Hussein M. A. Hussein ◽  
Ibrahim Ahmed ◽  
Emad Abouel Nasr ◽  
Ali Kamrani ◽  

Conventional design of jigs and fixtures has become unsuitable given the requirements of modern technology and complexity and diversity in the production with the rapid update of products. Computer-aided design (CAD) of jigs and fixtures is an effective solution in this direction. The current paper focuses on a computer-aided design of the traditional jigs and fixtures and developed a system containing tailor-made software, created using the Visual Basic programming language and installed on it the viewer screen to show the part. The developed system has been built by connecting Visual Basic programming language to the SolidWorks software on which the part is drawn and saved as STEP AP-203 file format, and the system reads and extracts the data from the STEP AP-203 file. Heuristic rules of feature recognition are pre-prepared for checking the extracted geometric data and deciding which data shape will represent the machining feature; then, the system provides the optimum design of the traditional jigs and fixtures for a group of hollow cylindrical parts that contain a group of cross-holes on the cylinder body, whether perpendicular or offset from the cylinder’s axis, (inclined or inclined offset, or blind or through, by applying pre-prepared heuristic rules for the design of traditional jigs and fixtures.

2021 ◽  
Vol 2021 ◽  
pp. 1-10
Fang Yuan ◽  
Yong Nie

With the rapid development of computer big data technology, online education in the form of online courses is increasingly becoming an important means of education. In order to objectively evaluate the teaching quality of online classroom, a teaching quality evaluation system based on facial feature recognition is proposed. The improved (MTCNN) multitask convolutional neural network is used to determine the face region, and then the eye and mouth regions are located according to the facial proportion relationship of the face. The light AlexNet classification based on Ghost module was used to detect the open and close state of eyes and mouth and combined with PERCLOS (percentage of eye closure) index values to achieve fatigue detection. Large range pose estimation from pitch, yaw, and roll angles can be achieved by easily locating facial feature angles. Finally, the fuzzy comprehensive evaluation method is used to evaluate students’ learning concentration. The simulation experiments are conducted, and the results show that the proposed system can objectively evaluate the teaching quality of online courses according to students' facial feature recognition.

Sign in / Sign up

Export Citation Format

Share Document