Machine Vision Applications Research Articles

A Machine Learning-Based Approach for Counting Blister Cards Within Drug Packages

Nowadays with the rapid development of technologies, machine vision has been used widely in various industries. The main applications of machine vision in industrial product lines are quality control (QC) and quality assurance (QA). The intelligent defects and anomalies recognition throughout the supply chain have come to be an integral part of quality control systems, in particular, in the food and pharmaceutical industries. In these industries, it is a legal requirement in manufacturing processes which can lead to minimizing the total number of defected products as well as maximizing the performance. In this paper, the machine vision has been utilized to monitor and control the proper packaging of drugs in pharmaceutical product lines. The main goal is counting the number of blister cards within a drug package. To tackle this problem, a new model based on object detection, feature extraction, and classification is proposed. Thanks to several strong approaches, such as the Haar cascade, HOG, ORG, Gabor wavelet, Radon transform, KNN, and SVM, and the accuracy over 88% is achieved in our experiments.

APPLICATION OF MACHINE VISION TECHNOLOGY IN ROAD DETECTION

Machine vision was first applied in industrial manufacturing field, and now it is also used in road detection, with the rapid development and continuous innovation of computer technology and digital image processing technology. This study provides a detailed description of the application of machine vision technology in detection of pavement crack, such as crack image acquisition, preprocessing (image de-noising and enhancement), segmentation, and recognition technology. Further the application of machine vision technology in pavement compactness and evenness was introduced. Finally, based on the application of machine vision in other aspects of road detection, hopefully, this study can provide a reference for method of road detection.

Development of six-DOF welding robot with machine vision

The application of machine vision to industrial robots is a hot topic in robot research nowadays. A welding robot with machine vision had been developed, which is convenient and flexible to reach the welding point with six degrees-of-freedom (DOF) manipulator, while the singularity of its movement trail is prevented, and the stability of the mechanism had been fully guaranteed. As the precise industry camera can capture the optical feature of the workpiece to reflect in the camera’s CCD lens, the workpiece is identified and located through a visual pattern recognition algorithm based on gray scale processing, on the gradient direction of edge pixel or on geometric element so that high-speed visual acquisition, image preprocessing, feature extraction and recognition, target location are integrated and hardware processing power is improved. Another task is to plan control strategy of control system, and the upper computer software is programmed in order that multi-axis motion trajectory is optimized and servo control is accomplished. Finally, prototype was developed and validation experiments show that the welding robot has high stability, high efficiency, high precision, even if welding joints are random and workpiece contour is irregular.

A boosted decision tree approach to shadow detection in scanning electron microscope (SEM) images for machine vision applications

Scanning electron microscopy is important across a wide range of machine vision applications, and the ability to detect shadows in images could provide an important tool for evaluating attributes of the surfaces being imaged, such as the presence of defects or particulate impurities. One example where the presence of shadows can be important is in the reconstruction of elevation maps from stereo-pair scanning electron microscopy (SEM) images. Shadows can both interfere with determination of matching points for stereoscopic calculations, and confuse shape-from-shading algorithms which rely on pixel intensity gradients to calculate surface slope, leading to inaccurate reconstructions. This paper describes a machine learning method for identifying locations in SEM images impacted by shadows, based on a training set of photographic images. The method could be useful as a means of identifying parts of images likely to suffer from reconstruction artifacts in shape-from-shading-based reconstructions, or as a tool for automated defect identification. The method uses a boosted decision tree machine learning approach to identify shadows based on the features of images. The method is illustrated with four different natural surfaces exhibiting a range of different types of shadow features, and an example is used to illustrate how the method can identify regions likely to be impacted by shadows in reconstructions.

Machine vision analysis on abnormal respiratory conditions of mice inhaling particles containing cadmium

Inhalable environmental toxicants can induce pulmonary malfunction resulting abnormal respiratory conditions. The traditional methods currently available to detect the respiratory condition of animals rely on differential pressure transducers and signal amplifiers. In comparison, current machine vision application requires little hardware. But it is unsuitable for respiratory condition tests of experimental animals reflecting respiratory toxicities of inhalable pollutants. In this study, we establish a new automatic method of machine vision analysis using a model that has mice inhaling aqueous aerosol with different concentrations of CdCl2 (0, 1, 3, 5 mM 2 h/day) for 7 days as simulant occupational exposure of inhalable Cd and analyze respiratory conditions such as respiratory rate, rhythm index, drive index and exchange index. Additionally, the models with different degrees of lung damage in mice are further tested and verified by the concentrations of cadmium accumulated in the lungs and the analyses on pulmonary porosity, fibrosis and inflammation. Machine vision analysis can identify the abnormal respiratory conditions of mice. Respiratory rate and rhythm index increase after exposure to cadmium. In the individuals with mild lung damage, respiratory drive index and exchange index in treatment group are higher than that in the control group, and in individuals with severe lung damage, these indices are similar to that of the control group. These abnormal respiratory conditions related to variable lung damage in mice demonstrate that the respiration is synchronously influenced by inhalable Cd and respiratory compensation according to normal physiological regulation, suggesting the present method is effective.

Digital Foveation: An Energy-Aware Machine Vision Framework

In machine vision applications, imaging systems and analysis algorithms are generally interdependent and energy intensive. We describe a machine vision energy minimization framework in which imaging hardware and vision algorithms are co-designed and tightly integrated. Digital foveation is inspired by the human vision system, which uses a spatially varying sensing architecture to generate oculomotory feedback and capture a series of high-resolution images using the densely sampling fovea. A multiround process with bidirectional information flow between camera hardware and analysis software optimizes energy consumption while preserving accuracy. By using existing hardware mechanisms, namely, row / column skipping, random access via readout circuitry, and frame preservation, digital foveation adapts to the chosen analysis algorithm. It aims to transmit and process only the necessary parts of the scene under consideration. This framework is general across a wide range of embedded machine vision applications and enables large improvements in energy efficiency. When evaluated for an embedded license plate recognition vision application, it reduces system energy consumption by 81.3% with at most 0.65% reduction in accuracy.

Cross-Modal Attentional Context Learning for RGB-D Object Detection.

Recognizing objects from simultaneously sensed photometric (RGB) and depth channels is a fundamental yet practical problem in many machine vision applications, such as robot grasping and autonomous driving. In this paper, we address this problem by developing a cross-modal attentional context (CMAC) learning framework, which enables the full exploitation of the context information from both RGB and depth data. Compared to existing RGB-D object detection frameworks, our approach has several appealing properties. First, it consists of an attention-based global context model for exploiting adaptive contextual information and incorporating this information into a region-based CNN (e.g., fast RCNN) framework to achieve improved object detection performance. Second, our CMAC framework further contains a fine-grained object part attention module to harness multiple discriminative object parts inside each possible object region for superior local feature representation. While greatly improving the accuracy of RGB-D object detection, the effective cross-modal information fusion as well as attentional context modeling in our proposed model provide an interpretable visualization scheme. Experimental results demonstrate that the proposed method significantly improves upon the state of the art on all public benchmarks.

A Variational Approach to Registration with Local Exponential Coordinates

AbstractWe identify a novel parameterization for the group of finite rotations (SO3), consisting of an atlas of exponential maps defined over local tangent planes, for the purpose of computing isometric transformations in registration problems that arise in machine vision applications. Together with a simple representation for translations, the resulting system of coordinates for rigid body motions is proper, free from singularities, is unrestricted in the magnitude of motions that can be represented and poses no difficulties in computer implementations despite their multi‐chart nature. Crucially, such a parameterization helps to admit varied types of data sets, to adopt data‐dependent error functionals for registration, seamlessly bridges correspondence and pose calculations, and inspires systematic variational procedures for computing optimal solutions. As a representative problem, we consider that of registering point clouds onto implicit surfaces without introducing any discretization of the latter. We derive coordinate‐free stationarity conditions, compute consistent linearizations, provide algorithms to compute optimal solutions and examine their performance with detailed examples. The algorithm generalizes naturally to registering curves and surfaces onto implicit manifolds, is directly adaptable to handle the familiar problem of pairwise registration of point clouds and allows for incorporating scale factors during registration.

Development of energy correction based multishot snapshot spectral imaging system

Abstract Spectrum of light reflected from an object is a fingerprint of its physical and chemical properties. In this paper, we proposed an energy correction based coded aperture snapshot spectral imaging (EC-CASSI) system to obtain the spatial/spectral data cube reflected from the object scene. Coded aperture was specifically designed in the system to efficiently compress input data in one shot. Based on the Two-step Iterative Shrinkage Thresholding (TwIST) algorithm, EC-TwIST was proposed with two-step energy correction to achieve higher accuracy in hyperspectral imaging reconstruction. Results showed that spatiospectral cube acquisition accuracy of EC-CASSI were 34.78% (differential relative spectral quadratic error evaluation) and 89.71% (entropy bias evaluation) higher than traditional CASSI system, respectively. This highly computation-relied spectrum obtainment method showed great potential in the application of machine vision and auto recognition.

Accurate Objects Detection Using Stereo Vision Sensor for Machine Vision Applications

This paper presents a new algorithm for object detection using a stereo camera system, which is applicable for machine vision applications. The propose algorithm has four stages which the first stage is matching cost computation. This step acquires the preliminary result using a pixel based differences method. Then, the second stage known as aggregation step uses a guided filter with fixed window support size. This filter is efficiently reduce the noise and increase the edge properties. After that, the optimization stage uses winner-takes-all (WTA) approach which selects the smallest matching differences value and normalized it to the disparity level. The last stage in the framework uses a bilateral filter, which is effectively further reduce the remaining noise on the disparity map. This map is two-dimensional mapping of the final result which contains information of object detection and locations. Based on the standard benchmarking stereo dataset, the proposed work produces good results and performs much better compared with some recently published methods.

Tunable fluidic lenses with high dioptric power

We report a complete theoretical model and supporting experimental results on the fabrication and characterization of macroscopic adaptive fluidic lenses with high dioptric power, tunable focal distance, and aperture shape. The lens is 17 mm wide and is made of an elastic polydimethylsiloxane (PDMS) polymer, which can adaptively restore accommodation distance within several cm according to the fluidic volume mechanically pumped in. Moreover, the lens can provide for magnification in the range of +25 diopter to +100 diopter with optical aberrations within a fraction of a wavelength, and overall lens weight of less than 2 g. The agreement between the non-linear theoretical model describing the elastic membrane deformation and the experimental results is apparent. We stress that these features make the proposed lenses appropriate for the low vision segment, as well as for applications in video magnifiers, camera zooms, telescope and microscopes objectives, and other machine vision applications where large magnification is required.

Memristor-CMOS Analog Coprocessor for Acceleration of High-Performance Computing Applications

Vector matrix multiplication computation underlies major applications in machine vision, deep learning, and scientific simulation. These applications require high computational speed and are run on platforms that are size, weight, and power constrained. With the transistor scaling coming to an end, existing digital hardware architectures will not be able to meet this increasing demand. Analog computation with its rich set of primitives and inherent parallel architecture can be faster, more efficient, and compact for some of these applications. One such primitive is a memristor-CMOS crossbar array-based vector matrix multiplication. In this article, we develop a memristor-CMOS analog coprocessor architecture that can handle floating-point computation. To demonstrate the working of the analog coprocessor at a system level, we use a new electronic design automation tool called PSpice Systems Option, which performs integrated cosimulation of MATLAB/Simulink and PSpice. It is shown that the analog coprocessor has a superior performance when compared to other processors, and a speedup of up to 12 × when compared to projected GPU performance is observed. Using the new PSpice Systems Option tool, various application simulations for image processing and solutions to partial differential equations are performed on the analog coprocessor model.<?enlrg 3pt?>

CLASSIFICATION OF COLOR CODED RESISTOR BASED ON POWER RATING

This paper describes the application of Machine Vision for classification of used color coded resistors into four different categories viz. 1/4 watt,1/2 watt,1 watt and 2watt. Physical dimension of the color coded resistor has been considered as an important feature that provides the power handling capability of that specific resistor. The blob measurement method has been employed for differentiating the resistors based on their power rating. This method could be further used for classification of lead broken resistors based on the material used for fabrication, which would be useful for recovery of the material during recycling process as well as dimension based classification of all types of Resistors

Evaluation of visual descriptors for painting categorisation

The ever increasing availability of digital data from the Arts and cultural heritage calls for efficient methods to organise, categorise, and retrieve such information in an effective and reliable way. In this context, painting classification has attracted significant research interest in recent years. In this work we address the problem of style classification, which involves determining the school, period and art movement to which a painting belongs. Notably, this job is peculiarly different from other machine vision applications – such as material, object and scene recognition – since the concept of ‘similarity’ is much more difficult to define in this case. For this specific task we evaluate, in this study, the effectiveness of an array of hand-designed visual descriptors against a set of feature extractors based on last-generation convolutional neural networks. We also investigate the effect of pre-processing methods such as image split and pyramidal decomposition. The experiments are based on the open-access Pandora dataset. The results show that pre-trained models can significantly outperform hand-designed descriptors with overall accuracy surpassing 67%. This represents an improvement on the state-of-the-art by ≈12 percentage points.

Gesture segmentation and classification using affine speed and energy

The characterization and analysis of hand gestures are challenging tasks with an important number of applications in human-computer interaction, machine vision and control, and medical gesture recognition. Specifically, several researchers have tried to develop objective evaluation methods of surgical skills for medical training. As a result, the adequate selection and extraction of similarities and differences between experts and novices have become an important challenge in this area. In particular, some of these works have shown that human movements performed during surgery can be described as a sequence of constant affine-speed trajectories. In this article, we will show that affine speed can be used to segment medical hand movements and present how the mechanical energy computed in the segment is analyzed to compare surgical skills. The position and orientation of the instrument end effectors are determined by six video photographic cameras. In addition, two laparoscopic instruments are capable of measuring simultaneously the forces and torques applied to the tool. Finally, we will report the results of these experiments and present a correlation between the mechanical energy values, dissipated during a procedure, and the surgical skills.

Adaptive Feedback in Local Coordinates for Real-time Vision-Based Motion Control Over Long Distances

We studied the differences in noise-effects, depth-correlated behavior of sensors, and errors caused by mapping between coordinate systems in robotic applications of machine vision. In particular, the highly range-dependent noise densities for semi-unknown object detection were considered. An equation is proposed to adapt estimation rules to dramatic changes of noise over longer distances. This algorithm also benefits the smooth feedback of wheels to overcome variable latencies of visual perception feedback. Experimental evaluation of the integrated system is presented with/without the algorithm to highlight its effectiveness.

Partial reset HDR image sensor with improved fixed pattern noise performance

An efficient method is presented to achieve 4T pixel dynamic range extension whilst keeping the fixed pattern noise (FPN) low. The method utilizes the floating diffusion node as a secondary photocharge integration node with a partial reset applied to the floating diffusion node. The output image signal is a recombination of the linear signal integrated in the photodiode node, and a compressed signal integrated in the floating diffusion node. The recombination algorithm suppresses the FPN stemming from the variable transfer gate potential barrier height. An image sensor array of 1344x1008 pixels was built in 0.11um technology. We demonstrate the preservation of low light sensitivity of 4T pixels, a dark noise of 3.5e-, and a dynamic range for the current sensor of 115 dB, with potential to greatly exceed 120-dB. Further, the FPN stemming from the transfer gate potential barrier height variation does not degrade the image; the pixel compressive response can be controlled on a per frame basis depending on the scene, and the HDR scene is captured during one frame. The sensor combines the properties that are crucial for automotive and other machine vision applications.

Disparity refinement process based on RANSAC plane fitting for machine vision applications

This paper presents a new disparity map refinement process for stereo matching algorithm and the refinement stage that will be implemented by partitioning the place or mask image and re-projected to the preliminary disparity images. This process is to refine the noise and sparse of initial disparity map from weakly textured. The plane fitting algorithm is using Random Sample Consensus. Two well-known stereo matching algorithms have been tested on this framework with different filtering techniques applied at disparity refinement stage. The framework is evaluated on three Middlebury datasets. The experimental results show that the proposed framework produces better-quality and more accurate than normal flow state-of-the-art stereo matching algorithms. The performance evaluations are based on standard image quality metrics i.e. structural similarity index measure, peak signal-to-noise ratio and mean square error.

An heuristic cloud based segmentation technique using edge and texture based two dimensional entropy

The Edge detection will localize the objects and their boundaries within an image which is a basis for various image analysis and the applications of machine vision. There are conventional approaches to edge detection which are expensive in terms of computation as each set of such operations are conducted for every pixel. In case of approaches that are conventional the time taken for computation will increase with that of the image size. The edge detection is used extensively in case of image segmentation of the medical images. An Ant Colony Optimization algorithm consists of many advantages such as parallelism, robustness and easy computation.The Glowworm Swarm Optimization (GSO) is that probabilistic technique which is used for finding the optimal paths that are connected completely in the guided search by means of using the information on brightness. The technique is also used for solving problems in computation that are reduced in finding. In case of the GSO algorithm where the insects move in a search space that is dictated probabilistically by using transition probabilities. In this work, a heuristic cloud based segmentation is performed using edge method and texture based 2-Dimensional entropy. The results have shown that this method proposed has achieved better performance.

Creating the Right Image

CMOS image sensors have developed to the point where they can be used in most machine vision applications.