Machine Vision Hardware Research Articles

Research on Online Measurement Method for Small Workpiece Circular Holes Based on Machine Vision

<p>In modern manufacturing, due to the narrow format, short size, and complex surface structure of small-sized workpieces, automatic detection of surface defects, dimensions, etc. on small-sized workpieces faces difficulties, making it difficult to meet the requirements of fully automated production lines. Especially, there is no good solution to the problem of high-precision automatic measurement of geometric features of circles and arcs on small-sized workpieces. This article takes the engine end cover circular hole thermal component as the research object, builds a hardware platform for image acquisition of small-sized workpieces based on machine vision, proposes an online size measurement method for small-sized circular holes, and realizes the measurement of the position size and aperture size of small-sized product shell workpieces. In order to achieve the above methods, this article builds a visual detection hardware system and describes a detailed machine vision hardware selection scheme. Then, the images collected by the system are preprocessed, including image denoising, image correction, and image feature extraction. Then, for the processed images, an improved Hough transform fitting method is used to obtain the online size of features. In order to visualize the results, the detection system and software are designed according to the actual application scenario. The overall scheme of this article is efficient and reasonable, and the detection result error is within a reasonable range.</p> <p> </p>

Wear monitoring solution for end mills using deep learning and mobile application

Mobile computing can effectively monitor tool life by establishing a relationship between periodically captured cutting tool images and the usefulness level of an end mill. This work presents an image-based deep-learning model to estimate end mill wear parameters and provide tool state feedback to the machine operator. The algorithm estimates the Remaining Useful Life (RUL) and tool wear state from the mobile camera images, viz. initial, intermediate, and worn. The operator captures cutting tool images at predefined intervals on the machine using a mobile camera, macro lens, and tripod arrangement. A pre-trained network, GoogLeNet, is employed for feature extraction, linear regression, and recognizing the tool wear status as RUL. A mobile application is developed to display the wear state and RUL for assisting machine operators in replacing/regrinding decisions. The accuracy and robustness of the proposed model are demonstrated using RMSE (Root Mean Square Error) and Correlation Coefficient (R2) metrics. A set of machining experiments are performed, and it has been shown that the developed module can capture wear states and RUL for end mills. The proposed solutions can be deployed effectively by manufacturing industries for obtaining tool wear information without significant investment in machine vision hardware and software.

Low Power Processors and Image Sensors for Vision-Based IoT Devices: A Review

With the advancements of the Internet of Things (IoT) technology, applications of battery powered machine vision based IoT devices is rapidly growing. While numerous research works are being conducted to develop low power hardware solutions for IoT devices, image capture and image processing remain high power demanding processes leading to a short battery life. However, the power consumption of the machine vision based IoT devices can be minimized by the careful optimization of the hardware components that are used is these devices. In this article, we present a review of low power machine vision hardware components for the IoT applications. A guide to selecting the optimum processors and image sensors for a given battery powered machine vision based IoT device is presented. Next, the factors that must be considered when selecting processors and image sensors for a given IoT application are discussed, and selection criteria for the processors and image sensors are established. Then, the current commercially available hardware components are reviewed in accordance with the established selection criteria. Finally, the research trends in the field of battery powered machine vision based IoT devices are discussed, and the potential future research directions in the field are presented.

Online System for Automatic Tropical Wood Recognition

There are more than 3000 wood species in tropical rainforests, each with their own unique wood anatomy that can be observed using naked eyes aided with a hand glass magnifier for species identification process. However, the number of certified personnel that have this acquired skills are limited due to lenghty training time. To overcome this problem, Center for Artificial Intelligence & Robotics (CAIRO) has developed an automatic wood recognition system known as KenalKayu that can recognize tropical wood species in less than a second, eliminating laborious manual human inspection which is exposed to human error and biasedness. KenalKayu integrates image acquisition, feature extraction, classifier and machine vision hardware such as camera, interfaces, PC and lighting. Grey level co-occurrence matrix (GLCM) is used for feature extraction. The features are trained in a back-propagation neural network (BPNN) for classification. This paper focusses more on the database development and the online testing of the wood recognition system. The accuracy of the online system is tested on different image quality such as image taken in low light condition, medium light condition or high light condition.

A Novel Auto-Sorting System for Chinese Cabbage Seeds.

This paper presents a novel machine vision-based auto-sorting system for Chinese cabbage seeds. The system comprises an inlet-outlet mechanism, machine vision hardware and software, and control system for sorting seed quality. The proposed method can estimate the shape, color, and textural features of seeds that are provided as input neurons of neural networks in order to classify seeds as “good” and “not good” (NG). The results show the accuracies of classification to be 91.53% and 88.95% for good and NG seeds, respectively. The experimental results indicate that Chinese cabbage seeds can be sorted efficiently using the developed system.

Review of Underwater Machine Vision Technology and Its Applications

AbstractThe characteristics of the underwater environment often present difficult challenges for machine vision technology. The images of inspected targets are distorted, blurred, and severely attenuated in an environment where lighting, visibility, and stability are generally not controllable. Therefore, underwater machine vision hardware and technology and their use in underwater robot/vehicle tasks such as ecological monitoring, map creation, and 3-D reconstruction provide ways to solve the described problems. The objective of this review is to highlight areas of research and development in the field of underwater machine vision that have made progress but have not been widely used for operational tasks. There are many potential underwater applications for this technology that could improve productivity and efficiency. To improve the accuracy of the underwater machine vision system, camera calibration and image preprocessing are still the focuses of research. If combined with other sensors, the system can better complete underwater tasks. There have been a number of initiatives in this area of study. Close collaboration with experts in other fields is also very important.

Аппаратно-программный комплекс компьютерного зрения для решения задач навигации наземных подвижных объектов

Аппаратно-программный комплекс компьютерного зрения для решения задач навигации наземных подвижных объектов

Variable field-of-view machine vision based row guidance of an agricultural robot

A novel variable field-of-view machine vision method was developed allowing an agricultural robot to navigate between rows in cornfields. The machine vision hardware consisted of a camera with pitch and yaw motion control. Guidance lines were detected using an image-processing algorithm, employing morphological features in a far, near and lateral field of view, and the robot was guided along these lines using fuzzy logic control.The method was tested while the vehicle successfully traveled through a distance of 30m towards the end of a crop row in three replications. To evaluate the guidance performance, RTK-GPS data were collected, showing a maximum guidance error of 15.8mm and stable navigational behavior.

Impact of object extraction methods on classification performance in surface inspection systems

In surface inspection applications, the main goal is to detect all areas which might contain defects or unacceptable imperfections, and to classify either every single ‘suspicious’ region or the investigated part as a whole. After an image is acquired by the machine vision hardware, all pixels that deviate from a pre-defined ‘ideal’ master image are set to a non-zero value, depending on the magnitude of deviation. This procedure leads to so-called “contrast images”, in which accumulations of bright pixels may appear, representing potentially defective areas. In this paper, various methods are presented for grouping these bright pixels together into meaningful objects, ranging from classical image processing techniques to machine-learning-based clustering approaches. One important issue here is to find reasonable groupings even for non-connected and widespread objects. In general, these objects correspond either to real faults or to pseudo-errors that do not affect the surface quality at all. The impact of different extraction methods on the accuracy of image classifiers will be studied. The classifiers are trained with feature vectors calculated for the extracted objects found in images labeled by the user and showing surfaces of production items. In our investigation artificially created contrast images will be considered as well as real ones recorded on-line at a CD imprint production and at an egg inspection system.

Machine vision developments

PurposeTo review recent developments in machine vision hardware and techniques, as revealed in the Image Processing and Optical Technology Exhibition, 2006.Design/methodology/approachDifferent sections of the paper examine new cameras and their interfacing, smart cameras with DSP processors, a powerful framegrabber, high intensity light emitting diode (LED) lighting, and developments in teraherz sources and detection.FindingsSmart cameras now exist with greater processing speed than a PC. A new generic programming interface aims to ease the incompatibilities of digital cameras. Water cooling has produced a very intense and long life LED light, and a novel goniophotometer has no moving parts. Rapid advances are occurring in the field of terahertz imaging.Originality/valueProvides an update on machine vision technology for scientists and engineers generally, and a source of contact information for specialists to follow up where necessary.

Machine Vision Monitoring for Automated Surface Finishing

A considerable amount of research has recently been performed on automated die and mold finishing systems. The research has tended to focus on the development of the finishing tool, the means of positioning and controlling the tool or efficient algorithms for moving the tool to achieve desired degrees of surface roughness. However, there has been relatively less effort to develop sensors suitable for providing the critical surface finish data necessary for any closed loop system. This paper presents two algorithms that, when coupled with machine vision hardware, are capable of providing surface texture information. The algorithms are developed and the results calibrated against a stylus profilometer. Tests have been conducted on mold cavity surfaces and the results evaluated against standard tactile means. The hardware has been incorporated with a computer controlled coordinate machine.