Papers by Keyword: Image Processing

Abstract: A majority of the research in Structural Health Monitoring focuses on detection of damage. This paper presents a method of imaging crack damage in an isotropic material using the Time Reversal imaging algorithm. Inputs for the algorithm are obtained via computational simulation of the propagation field of a crack in a medium under tone-burst excitation. The approach is similar to existing techniques such as Diffraction Tomography which makes use of the multi-static data matrix constructed using scatter field measurements from the computational simulation. Results indicate excellent reconstruction quality and accurate estimation of damage size.

Abstract: This article describes algorithm for patterns recognition of application for dual arm robot, android device and camera system. As first was create android application for getting information from computer via Bluetooth. Computer is using for image processing from external camera and then send image of dice to android devices and show on screen score of dice.

Abstract: The article deals with problems of image capture. There are presented methods of image processing through 2D and 3D camera systems. Section dedicated to image processing refers to the possibility of using different filters images, geometric and brightness transformations.

Abstract: The quality of product of manufacturing industries depends on dimension accurately and surface roughness quality. There are many types of surface defects and levels of surface roughness quality. Ironing process is one type of metal forming process, which aims to reduce the wall thickness of the cup-shaped or pipes products, thus increasing the height of the wall. Manually surface inspection procedures are very inadequate to ensure the surface in guaranteed quality. To ensure strict requirements of customers, the surface defect inspection based on image processing techniques has been found to be very effective and popular over the last two decades. The paper has been reviewed some papers based on image processing for defect detection. It has been tried to find some alternatives of useful methods for product surface defect detection of ironing process.

Abstract: The primary aim of the research, whose results are presented in this article, was to investigate the process of visual identification of objects. The objects recognised under this research were passenger seats for mass transit applications. The system presented in this paper is intended to be incorporated in the production process control system. The problem under analysis is based on the need to recognise (identify) an object located in the process area of the machine before proceeding with the process. Neural network was chosen as the adaptive element due to a high degree of similarity between the semi-products. The function of the network was to check if correct element had been fed. The input data fed to the network was the information from processing the photo images of these elements. The results of research as presented in this article are based on a limited set of elements and present the method used for solving the task of identification.

Abstract: Nowadays, machine engraving of photos into solid materials such as marble or granite is becoming very popular. Relatively cheap CNC machines are available. The problem is that high quality photos are essential to obtain good results. The first part of the paper describes a model of a CNC machine used for engraving and puts down the principles of image processing applied to poor quality photos in order to get the best results, as well as the fundamental image processing methods necessary for achieving satisfactory results when using an electromagnetic diamond percussion tool for engraving. The second part of the paper describes a very simple method of data coding and the algorithm of engraving tool movement for image engraving process by means of a control system based on ATmega16 microcontroller. The quality of the engraved images is comparable, or even better, than that of manually engraved images or images engraved by other competitive CNC machines.

Abstract: Nowadays, quality control becomes an important issue in semiconductor manufacturing industry. The rate of production with respect to time gives a lot of issues in the industry. In most semiconductor assemblies, a lot of defects generated from various processes in semiconductor wafer manufacturing need to be inspected manually using human experts and this process required full concentration of the operators. This human inspection procedure, however, is time consuming and highly subjective. In order to overcome this problem, implementation of machine vision will be the best solution. This paper presents defect segmentation of semiconductor wafer image based on colour features with k-Means clustering algorithm which can be adopted in machine vision system. In this work, the segmentation process is carried out in two stages. The first stage comprised of clustering the pixels in the image based on their colour and spatial features. Then the clustered pixels are merged to a specific number of regions. The proposed approach is being evaluated using defected wafer image. The experimental results show that it can be used to segment the defect correctly. By using this method, it is possible to increase the computational efficiency since it will avoid feature extraction for every pixel in the image.

Abstract: The study of bubble column reactors has its significance in applications such as multiphase reactors, aerators and in industrial waste-water treatment. Extensive works has been done in studying the hydrodynamics of a single gas bubble flowing through stationary liquid phase. The natural breakup of bubble during its motion has been studied in the past. In the Part I of the present work, hydrodynamics of an air bubble after its artificial splitting using a stainless steel mesh is experimentally studied using image processing and high speed photography. The significance of bubble splitting is that it increases the surface area of contact between stationery and moving fluid which in turn increases the rate of reaction desired during the process. The motion of the bubble is captured during its release and after splitting using High-Speed Camera. The velocity, area and diameter of the bubble before and after splitting are calculated by applying Image processing technique on the high speed photograph. The splitting of the bubble is found to vary with the superficial gaseous velocity. The splitting of bubbles into two bubbles of nearly equal size is considered and its hydrodynamic characteristics are studied.

Abstract: The paper presents a method for estimating the efficiency of the manual nesting process. By using the graphic file generated during the nesting process and using image processing techniques, the method allows the user to calculate the percentage of material used for manufacturing the parts. The method combines a manual approach - the user has to select some specific points on the graphic file with the mouse, with some image processing algorithms form Matlab software package.

Abstract: End milling is a key machining operation in industrial world, particularly in manufacturing of dies and similar products. Although, such products require high degree of surface roughness, milling operation is taken to be the enough for the cost wise if considering further finishing operations. Thus optimizing the cutting conditions to achieve the optimal surface roughness is becoming a vital issue. Several authors tackled this problem. In this paper the same case is investigated but with an advanced algorithm using regression and genetic methodology. The results obtained which ended by deducing a general equation combining the effect of various parameters on surface roughness highlighted the factors involved in achieving the surface roughness and proved to be good tool to predict the optimal cutting conditions.