Inspection Of Objects Research Articles

딥러닝 기반의 반도체 패키지 다이면 스크래치 검출 방법

The purpose of this paper is to evaluate the detection capability of deep learning-based surface scratches in semiconductor package dies and to determine whether deep learning can be employed as an objective inspection tool. To achieve this aim, detection capabilities are compared and evaluated based on a conventional image processing-based appearance inspection algorithm for examining equipment operating in the semiconductor package inspection process. In this study, a ResNet, which is the widely used CNN-based deep learning model, is used. The dataset was generated by collecting images of semiconductor packages being inspected in the semiconductor package appearance inspection process, which were cropped to focus on the parts necessary for the study. Three experiments were conducted for comparison. In the first experiment, the results of each detection using the Hough transform and ResNet-50 were compared. In the second experiment, the comparison was conducted in the same environment as in the first experiment by adding over-examination images. In the third experiment, the Hough transform algorithm and ResNet-50 were further compared by learning the over-examination images. Based on the results, an objective evaluation was conducted to determine whether deep learning could be applied to image processing-based machine inspection facilities.

The Use of Terrestrial and Maritime Autonomous Vehicles in Nonintrusive Object Inspection.

Traditional nonintrusive object inspection methods are complex or extremely expensive to apply in certain cases, such as inspection of enormous objects, underwater or maritime inspection, an unobtrusive inspection of a crowded place, etc. With the latest advances in robotics, autonomous self-driving vehicles could be applied for this task. The present study is devoted to a review of the existing and novel technologies and methods of using autonomous self-driving vehicles for nonintrusive object inspection. Both terrestrial and maritime self-driving vehicles, their typical construction, sets of sensors, and software algorithms used for implementing self-driving motion were analyzed. The standard types of sensors used for nonintrusive object inspection in security checks at the control points, which could be successfully implemented at self-driving vehicles, along with typical areas of implementation of such vehicles, were reviewed, analyzed, and classified.

Evolutionary Game Analysis of the Quality of Agricultural Products in Supply Chain

There are many factors affecting the quality and safety of agricultural products in the supply chain of agricultural products. In order to ensure the quality and safety of agricultural products, suppliers and processors need to take their own quality measures to ensure the quality of agricultural products. Quality inspection departments need to strictly supervise suppliers and processors to ensure the implementation of quality measures by both parties. Within the supply chain, the decisions of these three stakeholders are affected by the initial intention, the cost of quality measures, and the penalty amount of the quality inspection department. Outside the supply chain, they are affected by government regulation and consumer feedback. This paper takes the stakeholders in the agricultural product supply chain as the object, brings suppliers, processors, and quality inspection departments into the evolutionary game model, brings the factors that affect the decision-making of these three stakeholders into the model as parameters to analyze the stability of the model in different situations, and then analyzes the factors that affect the decision-making of stakeholders through mathematical simulation according to specific examples. The results show that the enthusiasm of stakeholders to ensure the quality of agricultural products is most affected by the initial intention of each other and the cost of quality measures. At the same time, the punishment of the quality inspection department, the feedback of consumers, and the supervision of the government also play a good role in promoting quality.

Object Detection Algorithm for Lingwu Long Jujubes Based on the Improved SSD

The detection of Lingwu long jujubes in a natural environment is of great significance for robotic picking. Therefore, a lightweight network of target detection based on the SSD (single shot multi-box detector) is presented to meet the requirements of a low computational complexity and enhanced precision. Traditional object detection methods need to load pre-trained weights, cannot change the network structure, and are limited by equipment resource conditions. This study proposes a lightweight SSD object detection method that can achieve a high detection accuracy without loading pre-trained weights and replace the Peleenet network with VGG16 as the trunk, which can acquire additional inputs from all of the previous layers and provide itself characteristic maps to all of the following layers. The coordinate attention module and global attention mechanism are added in the dense block, which boost models to more accurately locate and identify objects of interest. The Inceptionv2 module has been replaced in the first three additional layers of the SSD structure, so the multi-scale structure can enhance the capacity of the model to retrieve the characteristic messages. The output of each additional level is appended to the export of the sub-level through convolution and pooling operations in order to realize the integration of the image feature messages between the various levels. A dataset containing images of the Lingwu long jujubes was generated and augmented using pre-processing techniques such as noise reinforcement, light variation, and image spinning. To compare the performance of the modified SSD model to the original model, a number of experiments were conducted. The results indicate that the mAP (mean average precision) of the modified SSD algorithm for object inspection is 97.32%, the speed of detection is 41.15 fps, and the parameters are compressed to 30.37% of the original networks for the same Lingwu long jujubes datasets without loading pre-trained weights. The improved SSD target detection algorithm realizes a reduction in complexity, which is available for the lightweight adoption to a mobile platform and it provides references for the visual detection of robotic picking.

Laser Sensing System for Contactless Detection of Subsurface Defects in Concrete Tunnel Lining

Regular inspections of underground infrastructure are mandatory to maintain structural integrity and traffic safety. Standard procedures rely on visual inspection, complemented by testing for internal defects of the tunnel lining with the hammering method. This testing method uses manual excitation with a hammer and interpretation of the perceived sound by an engineer or technician. Result quality is based on the experience of the executing personal and thus highly subjective. The presented project aims at the automation of this procedure to enable a fast and objective inspection covering the whole tunnel lining. To achieve this, a new laser-based measurement system for detecting near surface defects in large concrete structures is developed. Surface vibrations are induced with plasma ignition and measured with a Doppler Vibrometer. The system has motorized mirror adjustment and can thus obtain fast-paced measurements on a grid. We present the experimental setup, the measurements performed on a concrete test block with artificial defects and the algorithms for automated defect detection. From the measured eigenfrequencies, we derive the size and location of defects. Computer vision techniques are adapted to make the system resilient against variable environmental conditions, like noise and temperature, or changes in construction material and age. These sophisticated signal processing methods are applied to extract relevant information from the measurements and are combined with topological data to provide service personnel with detailed maps for maintenance work.

Assessment of the nlmixr R package for population pharmacokinetic modeling: A metformin case study.

nlmixr offers first-order conditional estimation (FOCE), FOCE with interaction (FOCEi) and stochastic approximation estimation-maximisation (SAEM) to fit nonlinear mixed-effect models (NLMEM). We modelled metformin's pharmacokinetic data using nlmixr and investigated SAEM and FOCEi's performance with respect to bias and precision of parameter estimates, and robustness to initial estimates. Compartmental models were fitted. The final model was determined based on the objective function value and inspection of goodness-of-fit plots. The bias and precision of parameter estimates were compared between SAEM and FOCEi using stochastic simulations and estimations. For robustness, parameters were re-estimated as the initial estimates were perturbed 100 times and resultant changes evaluated. The absorption kinetics of metformin depend significantly on food status. Under the fasted state, the first-order absorption into the central compartment was preceded by zero-order infusion into the depot compartment, whereas for the fed state, the absorption into the depot was instantaneous followed by first-order absorption from depot into the central compartment. The means of relative mean estimation error (rMEE) ( ) and rRMSE ( ) were 0.48 and 0.35, respectively. All parameter estimates given by SAEM appeared to be narrowly distributed and were close to the true value used for simulation. In contrast, the distribution of estimates from FOCEi were skewed and more biased. When initial estimates were perturbed, FOCEi estimates were more biased and imprecise. nlmixr is reliable for NLMEM. SAEM was superior to FOCEi in terms of bias and precision, and more robust against initial estimate perturbations.

FIXATION AND EXAMINATION OF VOLUMETRIC TRACKS OF FOOTWEAR USING 3D SCANNING TECHNOLOGIES

The article is devoted to demonstrating the possibilities of using 3D scanning technologies to capture and study volumetric footprints. The author notes that traditional methods of visual fixation of such reflections (description, measurement, photography) are not informative enough, and the main method of object fixation (making plaster and other molds) is associated with the risk of significant damage to the fixed trace. At the same time, the use of contactless 3D scanners allows you to quickly, efficiently, and non-invasively create detailed digital 3D models of volumetric footprints. The author experimentally tested the fixation of shoe footprint in terms of inspection of the scene (at the place of its detection), scanning of the potential trace-forming object in terms of its separate inspection (in the laboratory), as well as a study of 3D models using laser 3D scanners Artec Leo and Artec Space Spider and Artec Studio 16 software. Possibilities of using Artec Studio 16 software to study and compare 3D models of forensically significant objects are demonstrated. The author’s experiments showed that with the use of 3D scanners with a scanning accuracy of up to 0.1 mm and the appropriate software, it is possible to scan and examine shoe prints with medium and large tread pattern elements, and using scanners with scanning accuracy up to 0.05 mm and more – fixation of traces left by shoes with more complex patterns of sole protector. The 3D models obtained in this way are accurate and detailed, accurately convey the size, shape, and color of scanned objects, and can be used for visual inspection and comparison of objects, and their visual demonstration (for example, in court). The level of detail of 3D models obtained with Artec Leo and Artec Space Spider allows comparative studies of shoe traces and the corresponding traceorming objects on their 3D models. Methods similar to those described in this article can be effectively used to capture and study other types of volumetric traces and their corresponding trace-forming objects: traces of tire treads and other parts of vehicles, breakage tools, bullet holes, etc. Key words: 3D scanning, 3D scanner, fixation, foot traces, comparative research.

Design of a compact RFQ linac for the transportable neutron source

A transportable accelerator-driven neutron source based on the Radio Frequency Quadrupole (RFQ) linac is developing at Xi’an Jiaotong University (X-TANS). It will be a promising tool for the non-destructive inspection of immovable objects. The neutron yield above 10 11n/s can be obtained via 7Li(p,n)7Be reaction with a proton beam energy of 2.5 MeV. To exactly guarantee the transportability of the neutron source, it is essential to achieve the light, compact and low-power consumption of RFQ linac. The operation frequency of 325 MHz is chosen considering the balance among cavity size, cavity consumption, and electrode aperture. In addition, compact dynamics scheme is adopted to achieve short RFQ by sacrificing a beam bunching efficiency. The beam dynamics design and optimization have been performed based on Parmteqm. Beam tracking results show that RFQ can accelerate the proton beam to 2.5 MeV with an electrode length of 2.6 m and an effective transmission efficiency of 93.8%. To enlarge the redundancy of the input beam mismatch and make the beam match easier from LEBT to RFQ, the electrode aperture is designed to shrink gradually in the RFQ entrance. In addition, we also performed the RF electromagnetic design and optimization based on the full-length 3D RFQ model. Multiphysics simulations of the cavity in different working conditions are investigated at last.

Method for remote control of radiation parameters of spacecraft based on X-ray fluorescence analysis

Existing international legal acts, instruments and procedures do not guarantee equal conditions for the exploration and use of outer space. There is a need for means of objective control of spacecraft carrying products with fissile materials. Inspection of such objects can be carried out by X-ray fluorescence methods. However, in the subject area under consideration, the use of such methods has been little studied. In this paper, a method is proposed for obtaining the spectra of X-ray fluorescence radiation of materials, the object under study based on the calculation of the spatial and energy characteristics of X-ray radiation. The X-ray fluorescence spectra of objects are obtained on the basis of the calculation of the spatial and energy characteristics of X-ray radiation according to the original geometric (mathematical) and simulation models developed by the authors. The calculations take into account the complex layered system of the object, taking into account the proportions of high-energy fluorescent radiation of the overlying layers. An original numerical experiment is proposed using a program that allows one to choose the projections of an object subjected to X-ray irradiation, the wavelength and intensity of the emitter parameters. Using the obtained spatial-energy distribution of radiation quanta and the physical properties of the radiation transmission medium, the problem of finding the coordinates and angles of intersection of tracks of quanta beams in each area of the object is solved. The result of software processing is displayed as the resulting spectrum. The obtained spectrum makes it possible to draw a conclusion about the chemical composition of the materials of the inspected object. The X-ray beam reaching the object is modeled as a spot with an area commensurate with the cross section of the device in the form of a selected geometric primitive — a square. The spot area of the incident photons is calculated from a predetermined divergence angle. On the basis of open literary sources, a physical model of an object with nuclear fissile materials W88 (USA) was chosen. The following characteristics of X-ray beams (subbeams) are accepted: a coherent beam of photons with a wavelength of 0.005 nm; beam scattering angle of 1 degree; emitter detector area of 4 m2. The resulting X-ray fluorescence spectrum was obtained which gives an idea of the chemical composition of the units and blocks of the apparatus and the object inside. The presence of specific products on the object board is confirmed by characteristic lines with normalized wavelengths indicating the presence of chemical elements belonging to the radioactive series. The results obtained can be used in the development of hardware and software for spacecraft devices that monitor the presence of fissile materials on board of the inspected vehicle.

A Universal Definition of Procedural Inspection as a Measure to Enforce the Proceedings in Administrative Cases: The Legal Essence and Problems of Legislative Regulation

The paper is devoted to the problems of applying such a security measure in cases of administrative offenses as inspection. The author explores the essence and potential of inspection as a procedural means of recording and documenting evidence in the framework of proceedings on administrative offenses. The current administrative procedure legislation provides for two types of inspections: inspection of the place of commission of an administrative offense (Article 28.1.1 of the Code of Administrative Offences of the Russian Federation) and inspection of premises, territories and objects located on them belonging to a legal entity or individual entrepreneur (Article 27.8 of the Code of Administrative Offences of the Russian Federation). At the same time, the content of these procedural actions, their procedural goals and grounds for application are not clearly defined in the law. In addition, the legal regulation of the inspection as a procedural action in the framework of an administrative case leaves open the question of all possible objects of such an inspection. Currently, the peculiarities of the legal regulation of the inspection do not allow the appropriate measure to be applied in all necessary cases. Moreover, its application carries high risks of recognizing the results obtained during the examination of the case as inadmissible evidence. As a solution to the above problems, it is proposed to consolidate in the administrative legislation a single unified concept — procedural inspection. Developing this proposal, the author provides the definition of a procedural inspection, develops general procedural requirements for its conduct. In addition, the paper substantiates the thesis about the need for special legal regulation and consolidation of special procedural requirements for conducting a procedural inspection of web pages (pages in information and telecommunications networks, including the Internet).

A Small UAV Optimized for Efficient Long-Range and VTOL Missions: An Experimental Tandem-Wing Quadplane Drone

Most types of Unmanned Aerial Vehicle (UAV, drone) missions requiring Vertical-Take-Off-and-Landing (VTOL) capability could benefit if a drone’s effective range could be extended. Example missions include Search-And-Rescue (SAR) operations, a remote inspection of distant objects, or parcel delivery. There are numerous research works on multi-rotor drones (e.g., quadcopters), fixed-wing drones, VTOL quadplanes, or tilt-motor/tilt-wing VTOLs. We propose a unique compact VTOL UAV optimized for long hover and long-range missions with great lifting capacity and manoeuvrability: a tandem-wing quadplane with fixed motors only. To the best of our knowledge, such a drone has not yet been researched. The drone was designed, built, and tested in flight. Construction details, its advantages, and issues are discussed in this research.

Active inspection of objects to detect possible damage and measure their volume using 3D reconstruction

Automated inspection processes offer the advantages of objectivity and efficiency for each test. For this reason, certain techniques for digital image processing and machine vision have been developed. However, it remains an open question due to the great variability in appearance and shape of the objects of study. In this work, we propose an inspection methodology that allows us to analyze objects for which we have developed an active inspection system to detect damage or malfunction. Its operation begins with the acquisition of images of the object using two cameras located in front of a rotating platform, with a controlled rotation, and projecting a laser line on the object. The following is the 3D reconstruction of a section of the object, to then make a comparison using a 3D descriptor and distance analysis between points of the 3D test model and the 3D model generated with a base version of the same object without damage, which was stored in a database. If any damage is detected, the inspection is continued until the entire damaged section is completed and its volume is calculated. The main contributions of our work are summarized as: (i) Efficient combination of machine vision algorithms for volume estimation, (ii) Active inspection in the automatic search for surface damage on objects, and (iii) Measurement of the damage volume detected in metric units , which can be subtractive or additive damage. To validate our proposal, we have applied this methodology to different objects, thus being able to estimate the volume of damage (if any). Our results are promising, as we achieved a Mean Absolute Percentage Error, MAPE=2.78%.

Feature detection network-based correction method for accurate nano-tomography reconstruction.

Driven by the development of advanced x-ray optics such as Fresnel zone plates, nano-resolution full-field transmission x-ray microscopy (Nano-CT) has become a powerful technique for the non-destructive volumetric inspection of objects and has long been developed at different synchrotron radiation facilities. However, Nano-CT data are often associated with random sample jitter because of the drift or radial/axial error motion of the rotation stage during measurement. Without a proper sample jitter correction process prior to reconstruction, the use of Nano-CT in providing accurate 3D structure information for samples is almost impossible. In this paper, to realize accurate 3D reconstruction for Nano-CT, a correction method based on a feature detection neural network, which can automatically extract target features from a projective image and precisely correct sample jitter errors, is proposed, thereby resulting in high-quality nanoscale 3D reconstruction. Compared with other feature detection methods, even if the target feature is overlapped by other high-density materials or impurities, the proposed Nano-CT correction method still acquires sub-pixel accuracy in geometrical correction and is more suitable for Nano-CT reconstruction because of its universal and faster correction speed. The simulated and experimental datasets demonstrated the reliability and validity of the proposed Nano-CT correction method.

Development of automatic detection model for stem-end rots of ‘Hass’ avocado fruit using X-ray imaging and image processing

Postharvest qualitative avocado losses attributable to stem-end rot are serious problems in various main production areas. Postharvest inspection of avocadoes is still commonly performed by hand; this process depends upon only slight differences in fruit appearance or firmness, and it can lead to misclassification. The development of a non-destructive stem-end rot detection technique would help improve objective inspection and reduce misclassification. Therefore, this study proposes an X-ray machine vision approach to automatically detect stem-end rot in avocados. Four-hundred-and-twenty-two X-ray images of fully ripened ‘Hass’ avocadoes were obtained to develop and validate an automatic classification model. The stem-end rot features visible in the X-ray images were extracted using two image processing operations: image segmentation and fruit-outline evaluation. Then, an extremely randomized trees model classified rot development into two classes (< 20% and ≥ 20%) using two features extracted from the X-ray image. The overall classification accuracy was 0.89 with an area under the curve of 0.93 and a recall of 0.89, which suggests the possibility of performing stem-end rot inspections via X-ray imaging and image processing.

Remote inspection method for water conveyance tunnel based on single-beam scanning sonar

Water conveyance tunnels perform an important role in water-diversion projects. With a water conveyance tunnel transferring water, the conflict between urban crowded ecology and agricultural water can be relieved. Research shows that there is a significant need for a more objective and accurate inspection of water conveyance tunnels during operation. Hence, we investigate the use of a single-beam scanning sonar to detect a water conveyance tunnel and generate a three-dimensional point cloud model, which can be used to make objective predictions for surface defects. First, a cobweb-like grid matrix was proposed to grid discrete points to establish the topological relationship of the point cloud. Second, a denoising method fusing acoustic properties and intensity was used to filter the noise in the acoustic point cloud. Third, an optimal echo extraction technology based on a second-order differential model was proposed to obtain a high-precision point cloud of the tunnel. Experiments show that the proposed denoising method and optimal echo extraction method are superior to traditional algorithms commonly used in ground point clouds, with good results achieved at real test sites.

Automatic anomaly detection from X-ray images based on autoencoders

ABSTRACT Radiography is a common imaging method for non-destructive object inspection. Processing radiographs to detect defects is, however, challenging, especially if the defect type and location are unknown. In other fields, autoencoders (AE) have been largely studied for anomaly detection. While their performance for defect detection has shown promise, many solutions are not demonstrated to be defect-independent. In this work, we introduce a radiography-based anomaly detection method that first computes the AE prediction error from an input radiograph to enhance the detectability of defects. Subsequently, the probability of the imaged object being defective is predicted by a convolutional classification network. Results on simulated data demonstrate that this method succeeds in distinguishing between defective and non-defective objects, with accuracy (which calculates how often the predictions are equal to the labels) and precision (which is the ratio between the true positives and the total predicted positive) , even for defects unseen during the training stage. Our classification network showed high robustness to noise (with signal-to-noise-ratio (SNR) of 20 dB), and accuracy on noisy data and precision . Since our network workflow does not require image registration for defect detection, the proposed solution is therefore independent of the orientation and position of the object during the scan.

A Proposal of Template Matching Based on Pixel Selection for Constraint Rotated Patterns

Image matching technology is an important technology used for positioning and inspection of objects in the factory automation field. One of the most common recognition methods is template matching. However, this method requires matching a large number of templates in order to detect rotated objects in an image. Therefore, it has a problem of computation time. We propose to match only a small number of templates by assigning a rotation variation tolerance to the templates. The rotation variation tolerance is given by selecting pixels from the templates. We propose a new method of matching with only a few templates. Experimental results showed the method achieved a 91.9% recognition rate and a 288 msec processing time.

Application research and development of objective examination of olfactory function

The sense of smell is one of the five most primitive human sensory functions, and it plays a very important role in our daily lives. Despite numerous methods for evaluating olfactory function, there is still a lack of standardization of olfactory tests and the results are often inconsistent. Furthermore, the related research on objective evaluation started relatively late. Along with the deciphering of the olfactory pathway, the technical level of olfactory objective inspection has been greatly improved and significant progress has also been made in terms of clinical application, such as: olfactory pathway MRI and fMRI imaging, OERPs, BEAM for various olfactory disorders and early diagnosis of neurodegenerative disorders, as well as related research based on bionic olfactory sensing technology. This article mainly introduces the recent research progress of several commonly used objective olfactory tests and provides reference for more accurate evaluation of olfactory function.

Research on defect type diagnosis of composite insulator based on infrared detection technology

Composite insulator is widely used in Hainan area of China because of its lightweight, small volume, good anti pollution flashover performance and many other advantages. Aiming at the problem of how to correctly judge the aging state of composite insulators under live conditions in Hainan’s high humidity environment, this paper analyzes the main causes of composite insulator heating and the heating position and temperature rise of composite insulator under different defect types by building equivalent circuit, and puts forward several criteria for distinguishing the defect types of composite insulator by infrared detection technology; taking the abnormal temperature rise of 220kV returned composite insulator in Hainan as an example, the defect type of the composite insulator is analyzed by infrared detection technology, and the correctness of the infrared results is verified by the appearance inspection and anatomical analysis of the test object. The results show that the defect that the temperature rise is between 1-5°C and only appears among the fittings and the first umbrella is caused by the aging, moisture absorption or surface pollution of the umbrella cover; the defect that the temperature rise is above 5°C and the range extends from the high-voltage end sheath to several umbrella skirts is caused by the internal deterioration of the insulator. In conclusion, the conclusion provides a guarantee for the safe operation of power grid and has important reference value for the study of on-line monitoring of composite insulators.

MSFYOLO: Feature fusion-based detection for small objects

At present, the effect of object detection algorithm in small object detection is very poor, mainly because the low-level network lacks semantic information and the characteristic information expressed by small object inspection data is very lack. In view of the above difficulties, this paper proposes a small object detection algorithm based on multi-scale feature fusion. By learning shallow features at the shallow level and deep features at the deep level, the proposed multi-scale feature learning scheme focuses on the fusion of concrete features and abstract features. It constructs object detector (MSFYOLO) based on multi-scale deep feature learning network and considers the relationship between a single object and local environment. Combining global information with local information, the feature pyramid is constructed by fusing different depth feature layers in the network. In addition, this paper also proposes a new feature extraction network (CourNet), through the way of feature visualization compared with the mainstream backbone network, the network can better express the small object feature information. The proposed algorithm is valuated on the MS COCO and achieved leading performance with 11.7% improvement in FPS, 17.0% improvement in AP, 81.0% improvement in ARS, and 23.3% reduction in computational FPLOs compared to YOLOv3. This study shows that the combination of global information and local information is helpful to detect the expression of small objects in different illumination. MSFYOLO uses CourNet as the backbone network, which has high efficiency and a good balance between accuracy and speed.