A New Tool for Automatic Classification of Microstructure Based on Backpropagation Artificial Neural Network

This paper presents a comparative analysis between two automatic systems to segment and quantify microstructures of nodular cast iron, malleable cast iron, and gray cast iron, in images. The compared computational systems are the SVRNA system (Computational Vision System based on an Artificial Neural Network for Microstructure Segmentation and Quantification), developed during this work and here presented, that uses an artificial neural network based on the backpropagation algorithm to segment and quantify microstructures in metallic materials, and the Image Pro-Plus, a common tool used for the same purpose. For results comparison 60 samples of cast iron had been considered and analyzed, and the results of our SVRNA system are very similar to the ones obtained by visual human inspection. In fact, the SVRNA system has segmented efficiently and automatically the microstructures of the materials in analysis, what has not always occurred with the Image Pro-Plus. Thus, we could conclude that our SVRNA system is a valid and adequate option for researchers, engineers, specialists and others of Material Sciences field accomplish microstructural analysis from images in a fully automatic and efficient manner

Nondestructive Testing and Evaluation Fo rP ee SVRNA System A New Tool for Automatic Classification of Microstructure Based on Backpropagation Artificial Neural Network Manuscript ID: Manuscript Type: Complete List of Authors: GNTE-2007-0002 Original Article 15-Dec-2007 ev Date Submitted by the Author: Nondestructive Testing and Evaluation rR Journal: w ie Albuquerque, Victor; Universidade Porto Cortez, Paulo; Universidade Federal Ceará, Departamento de Engenharia de Teleinformática Alexandria, Auzuir; Universidade Federal Ceará, Departamento de Engenharia de Teleinformática Tavares, João; Universidade Porto artificial neural networks, image processing and analysis, image segmentation and quantification, Micrstructure classification ly On Keywords: URL: http://mc.manuscriptcentral.com/gnte Page of 58 SVRNA system – a new tool for automatic classification of microstructure based on backpropagation artificial neural network V H C de Albuquerque 1*, P C Cortez2, A R de Alexandria2, João Manuel R S rP Tavares1 Fo Instituto de Engenharia Mecânica e Gestão Industrial (INEGI), Faculdade de Engenharia da Universidade Porto (FEUP), Departamento de Mecânica e Gestão ee Industrial (DEMEGI), Rua Dr Roberto Frias, S/N, 4200 - 465 Porto – Portugal rR Universidade Federal Ceará (UFC), Departamento de Engenharia de Teleinformática, Avenida Mister Hull, S/N – Pici, CEP 60455.970 - Fortaleza – Ceará, ev Brasil w ie (Received 15 Dcember 2007; final version received xx August 2008) ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation Abstract This paper presents a comparative analysis between two automatic systems to segment and quantify microstructures of nodular cast iron, malleable cast iron, and gray cast iron, in images The compared computational systems are the SVRNA system (Computational Vision System based on an Artificial Neural Network for Microstructure Segmentation and Fo Quantification), developed during this work and here presented, that uses an artificial rP neural network based on the backpropagation algorithm to segment and quantify microstructures in metallic materials, and the Image Pro-Plus, a common tool used for the ee same purpose For results comparison 60 samples of cast iron had been considered and analyzed, and the results of our SVRNA system are very similar to the ones obtained by rR visual human inspection In fact, the SVRNA system has segmented efficiently and automatically the microstructures of the materials in analysis, what has not always occurred ev with the Image Pro-Plus Thus, we could conclude that our SVRNA system is a valid and ie adequate option for researchers, engineers, specialists and others of Material Sciences field w accomplish microstructural analysis from images in a fully automatic and efficient manner ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Key-words: Artificial neural networks; Image processing and analysis; Image segmentation and quantification; Microstructure; Multilayer Perceptron; Materials Science *Corresponding author Email: dce06017@fe.up.pt URL: http://mc.manuscriptcentral.com/gnte Page of 58 Page of 58 Introduction One of the greatest existing challenges in the development of machines and equipments is the conception of systems with intelligent capacities, so as they can consider problems in an analog way to humans Thus, various advances have occur in the domain of Artificial Intelligence, which has as main objective the research of new algorithms and Fo technological solutions for the construction of systems with intelligent capacities However, rP this is a very difficult and complex task, as there are innumerable tasks that humans commonly and naturally as, for example, visualize, hear, walking and speech, but that ee are not trivial process to implement in computational systems Among the tasks above cited, the vision has deserved special attention of the rR scientific community; in particular, because of the considerable number of existing ev applications and its major importance in our society [1] Initially derived from Artificial Intelligence field, Computational Vision became a distinct research area that searches to ie build computational systems able to perform some visual information analysis and w interpretation, which can be capable to help humans to execute some usual tasks with higher speed and precision [2] This new area uses, between others, techniques from ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation Artificial Intelligence, Digital Signal Processing and Pattern Recognition fields [3] Artificial Neural Networks (ANN) is one of the techniques used in Artificial Intelligence and Pattern Recognition fields In particularly, they have been used in applications that involve Pattern Recognition with high degree of parallelism, considerable classification speed and important capacity to learn through examples [4] URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation Artificial Neural Networks have been used in some domains of Materials Science as well Some examples that can be found are: in welding control [5], to obtain the relations between process parameters and correlations in Charpy impact tests [6], to get the composition of models for ceramic matrices [6] , in the modeling of alloy elements [7,8], prediction of welding parameters in pipeline welding [9], modelling of microstructure and Fo mechanical properties of steel [10], model based on deformation mechanism of titanium rP alloy in hot forming [11], prediction of properties of austempered ductile iron [12], predict the carbon content and the grain size of carbon steels [13], models for predicting flow stress ee and microstructure evolution of a hydrogenized titanium alloy [14], among others In this work, the SVRNA system developed in the scope of this work is present rR This system is able to accomplish the quantification of microstructures in metallic materials from images Also, in this work, the SVRNA system is compared with the Image Pro-Plus ev system, which is a image analysis software for fluorescence imaging, materials imaging, ie and various other scientific areas as, for example, biomedical engineer and industrial w applications [15] For this comparison, we used experimental samples of nodular, malleable and gray cast iron These materials were selected manly because of their large use in ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 industry, as in, for example, machine base structures, lamination cylinders, main bodies of valves and pumps, gear elements, among others Materials and methods In this work, the materials considered to be analyzed are nodular cast iron, malleable cast iron and gray cast iron, which present microstructures with different shapes URL: http://mc.manuscriptcentral.com/gnte Page of 58 Page of 58 Thereafter, we need the application of computational techniques to automatically segment and quantify the constituents of the cast irons considered that are presented in images acquired for that purpose To accomplish this, it has developed the already referred SVRNA system, which is based on an Artificial Neural Network (ANN) As the nervous system, ANN not possess discrete functional components Fo However, for a determined task, there are responsible regions with high degree of rP redundancy and parallelism, which turn the computational systems based on ANN usually robust and fast ee The ANN are composed by a set of perceptrons, being the perceptron a model of a nervous cell (neuron) Such element is the most basic one that we can find on an ANN [16] rR Perceptron networks are neural networks of a unique layer that recognize only linear separable patterns; in general, these networks are only used in simple tasks ev For the solution of more complex problems, the networks MLP (MultiLayers ie Perceptron) are used, whose pattern, in general, are not linear separable Usually, these w networks consist of an input layer, several hidden layers and an output layer The same nets possess two distinct phases: training and execution In this kind of neural networks is ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation widely used the backpropagation algorithm, as well as are its variants [17] The learning rule of this kind of networks is considerably more complex than the rule used for simple perceptron networks This learning is also known as "supervised", because it is necessary to know previously the correct pattern so that the training is carried through successfully, needing an adequate operator for that [17] URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation 2.1 Cast Iron Cast irons are composed by iron and carbon and possess contents of this last one Fo between 2.14 and 6.70% In fact, the majority of cast irons contain between 3.0 and 4.5% of rP carbon [18] Beside carbon, the cast irons present significant silicon contents, and, therefore, some authors as, for example, Callister [18], Chiaverini [19] and Raabe et al ee [20], consider the cast iron as an iron alloy, with carbon and silicon Another characteristic of the cast irons is associated with the form that carbon presents in its microstructure For rR example, in gray cast irons, all the carbon is presented practically free in the form of lamellar graphite or in veins, while that in the white cast iron, the carbon presented is ev arranged in the form of cementite (Fe3C) [21] ie In accordance with the carbon content, cast irons can be classified in hipoeutectics, w eutectics or hipereutectics, as it is shown in Figure When eutectic cast iron is solidified, just below of point G (see Figure 1), it happens the formation of a microstructure with ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 cementite and austenite globules, called ledeburite I Carrying on the cooling process, below the 727 ºC line, the austenite transforms itself into pearlite globules on the cementite, forming ledeburite II [22], that was particularly considered in this work Line I of the iron-carbon diagram shown in Figure 1, specifies the position of a hipoeutectic cast iron or, using other words, with carbon content inferior than 4.3% Line II, in turn, specifies a hipereutetic cast iron, that presents carbon content superior than 4.3% [18] URL: http://mc.manuscriptcentral.com/gnte Page of 58 Page of 58 SVRNA system Our SVRNA system was implemented in C++ using the environment Builder from Borland, runs in Microsoft Windows XP platforms and, as already referred, uses an Fo artificial neural network, trained using the backpropagation algorithm, to accomplish the rP segmentation and quantification of microstructures in metallographic images More specifically, an artificial neural network composed by 42 neurons distributed in two layers ee is used The neuron distribution along the ANN layers (topology) corresponds to: inputs, 30 neurons in the hidden layer and neurons in the output layer The inputs for the neural rR network are the R (red), G (green) and B (blue) components, or the gray level instead, of pixels selected in the input image The output of the same network is the indication of the ev region, which must be associated to each pixel considered, through the indication of the ie pseudocolor that must be attributed in its labeling w As previously described, the neural network integrated in our SVRNA system is used to segment the original images acquired for the analysis of microstructures, being their ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation classification and quantification made from the pseudocolors associated to the same ones The training of the network used is made considering a reduce sample set of pixels from each microstructure to consider, attributing to each set a pseudocolor that is used as the label of the associated microstructure Thus, on the training of the neural network employed are used example sets of components R, G, B, or of correspondent gray levels, and the associated pseudocolors (numbered from to 8) URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation The developed system interacts with its user in order to collect the pixels of the selected example sets, considering for that the computer mouse These example sets are then used in the training phase of the neural network and to obtain the correspondent microstructures classification With an adequate training of the network used, the same network can be used in the classification and quantification of similar images of the Fo identical materials rP The following computational modules constitute our SVRNA system: training, manipulation of records, execution and output The training module is responsible, as its ee proper name indicates, for the learning phase of the neural network used, being the backpropagation algorithm considered for that of supervised type The functions of the rR manipulation of records module are: the reading and generation of new records The execution module of SVRNA is responsible for the segmentation of the microstructures ev present in the original input image and, consequently, for its classification This module ie just can be used after the training process of the neural network used, and it is initiated with w the opening of the record associated to an input image and it is finished with the display of the segmented image and the obtained numerical results, using for that an area chart with ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 the percentage of each classified microstructure The segmented images and the area chart are the outputs of the segmentation process carried through the neural network used, and can be visualized and saved for posterior consideration 3.1 SVRNA usage URL: http://mc.manuscriptcentral.com/gnte Page of 58 Page of 58 To segment and quantify the microstructures present in material samples using our SVRNA system, is necessary to carry through a process of samples preparation: metallography In other words, we need to section, countersink, sand, polish and etch each sample to the be analyzed After that, the images of each material sample are acquired using a digital camera direct-coupled to an optical microscope Fo In Figure 2, is presented the interface of SVRNA system In this image, we can rP visualize an original image of a gray cast iron and the corresponding segmented image obtained using our SVRNA system ee The use of our SVRNA system is very practical and intuitive, and it consists in the following First, the user opens the original image to be analyzed Then, the user resets the rR neural network and afterward, for each microstructure to be segmented, selects the desired pseudocolors and initiates the selection process of some pixels that should be considered by ev the network for the posterior segmentation of that microstructure For example, in Figure 2, ie the yellow and blue points represent the pixels selected to segment the pearlite and graphite w microstructure of a gray cast iron Whenever this selection process is finish, the user starts the training of the network used, using for this the data selected After the network training, ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation the system is ready to automatic segments and quantifies the input image The output results of this procedure are a segmented image, which can be saved for posterior consideration, and the numerical values of the quantification obtained, that can be also saved and exported to others applications, like the Microsoft Excel It is important to note that the training of the neural network used is only necessary to be done once for each type of sample to be analyzed In addition, the SVRNA system lets the pseudocolors adjust of the segmentation obtained, that can be used to improve the URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation Fo rR ee rP Figure 7: Original image of the AISI 1020 steel, a), and after the segmentation done using the SVRNA system, b) 50x33mm (300 x 300 DPI) w ie ev ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 URL: http://mc.manuscriptcentral.com/gnte Page 44 of 58 Page 45 of 58 Fo rR ee rP Figure 7: Original image of the AISI 1020 steel, a), and after the segmentation done using the SVRNA system, b) 50x33mm (300 x 300 DPI) w ie ev ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation Fo rR ee rP Figure 8: Original image of AISI 1045 steel, a), and after the segmentation done using the SVRNA system, b) 50x33mm (300 x 300 DPI) w ie ev ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 URL: http://mc.manuscriptcentral.com/gnte Page 46 of 58 Page 47 of 58 Fo rR ee rP Figure 8: Original image of AISI 1045 steel, a), and after the segmentation done using the SVRNA system, b) 50x33mm (300 x 300 DPI) w ie ev ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation Fo rR ee rP Figure 9: Original images of inclusions, a) and c), segmented images resulting using the SVRNA system, b) and d) 50x33mm (300 x 300 DPI) w ie ev ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 URL: http://mc.manuscriptcentral.com/gnte Page 48 of 58 Page 49 of 58 Fo rR ee rP Figure 9: Original images of inclusions, a) and c), segmented images resulting using the SVRNA system, b) and d) 50x33mm (300 x 300 DPI) w ie ev ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation Fo rR ee rP Figure 9: Original images of inclusions, a) and c), segmented images resulting using the SVRNA system, b) and d) 50x33mm (300 x 300 DPI) w ie ev ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 URL: http://mc.manuscriptcentral.com/gnte Page 50 of 58 Page 51 of 58 Fo rR ee rP Figure 9: Original images of inclusions, a) and c), segmented images resulting using the SVRNA system, b) and d) 50x33mm (300 x 300 DPI) w ie ev ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation Table 1: Results obtained using the SVRNA and Image Pro-Plus systems on samples of nodular cast iron Nodular cast iron SVRNA Image Pro-Plus Samples r Fo Graphite (%) Pearlite (%) Graphite (%) Pearlite (%) Sample1 11.51 88.49 17.52 82.48 Sample 13.36 86.64 19.91 80.09 13.19 86.81 18.05 81.95 13.46 86.54 17.63 82.37 Sample 12.46 87.54 16.98 83.02 Sample 11.79 88.21 17.07 82.93 Sample 14.58 85.42 18.38 81.62 Sample 12.50 8.50 17.79 82.21 Sample 14.02 85.98 21.12 78.88 Sample 10 13.30 86.70 Sample 11 9.08 Sample12 Sample er Sample Pe w vie Re 18.12 81.88 90.92 12.44 87.56 9.25 90.75 11.85 88.15 Sample 13 11.56 88.44 14.35 85.65 Sample 14 9.24 90.76 11.39 Sample 15 10.22 89.78 12.82 87.18 Sample 16 9.89 90.11 12.74 87.26 Sample17 8.31 91.69 11.30 88.7 ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 52 of 58 URL: http://mc.manuscriptcentral.com/gnte 88.61 Page 53 of 58 Sample 18 7.66 92.34 10.03 89.97 Sample 19 14.66 85.34 11.37 88.63 Sample 20 21.73 78.27 27.77 72.23 Average 12.09 87.91 15.93 84.07 r Fo er Pe w vie Re ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation URL: http://mc.manuscriptcentral.com/gnte Nondestructive Testing and Evaluation Table 2: Average results obtained using the manual quantification method, SVRNA and Image Pro-Plus system, on samples of nodular cast iron Method Graphite (%) Pearlite/Ferrite (%) Manual 24.13 75.87 SVRNA 12.09 87.91 15.93 84.07 r Fo Image Pro-Plus er Pe w vie Re ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 54 of 58 URL: http://mc.manuscriptcentral.com/gnte Page 55 of 58 Table 3: Results obtained using the SVRNA and Image Pro-Plus systems on samples of gray cast iron Gray cast iron SVRNA r Fo Image Pro-Plus Samples Graphite (%) Pearlite (%) Graphite (%) Pearlite (%) Sample1 12.21 87.79 12.31 87.69 Sample2 6.80 7.20 92.80 8.56 91.44 10.70 89.30 93.20 Sample4 6.80 93.20 8.36 91.64 Sample5 8.03 91.97 10.96 89.04 Sample6 5.02 94.98 6.91 93.09 Sample7 8.73 91.27 11.21 88.79 Sample8 9.15 90.85 11.00 89.00 Sample9 7.90 92.10 Sample10 6.67 93.33 Sample11 40.41 59.59 46.85 53.15 Sample12 7.47 92.53 10.66 89.34 Sample13 9.83 90.17 12.77 Sample14 7.52 92.48 10.52 89.48 Sample15 21.69 78.31 15.40 84.60 Sample16 8.79 91.21 9.70 90.30 Sample 17 7.76 92.24 10.21 89.79 Sample3 er Pe w vie Re 9.82 90.18 8.35 91.65 ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation URL: http://mc.manuscriptcentral.com/gnte 87.23 Nondestructive Testing and Evaluation Sample18 6.77 93.23 9.28 90.72 Sample19 7.64 90.28 8.21 91.79 Sample 20 7.78 92.22 11.35 88.65 Average 10.28 89.62 12.09 87.91 r Fo er Pe w vie Re ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 56 of 58 URL: http://mc.manuscriptcentral.com/gnte Page 57 of 58 Table 4: Results obtained using the SVRNA and Image Pro-Plus systems on samples of malleable cast iron Malleable cast iron r Fo SVRNA Image Pro-Plus Graphite (%) Pearlite (%) Graphite (%) Pearlite (%) Samples Sample 81.05 31.65 68.35 15.71 84.29 24.48 75.52 14.96 85.04 28.46 71.54 Sample 14.00 86.00 29.16 70.84 Sample 15.16 84.84 24.08 75.92 Sample 16.07 83.93 21.11 78.89 Sample 19.11 80.89 30.55 69.45 Sample 19.22 80.78 27.58 72.42 Sample 15.64 84.36 Sample 10 17.64 Sample 11 Sample er Sample 18.95 Pe 76.11 82.36 31.23 68.77 11.84 88.16 18.23 81.77 Sample 12 12.04 87.96 21.12 78.88 Sample 13 13.72 86.28 22.62 Sample 14 14.06 85.94 22.24 77.76 Sample 15 11.83 88.17 17.15 82.85 Sample 16 11.40 88.60 16.16 83.84 On 23.89 ly w vie Re 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Nondestructive Testing and Evaluation URL: http://mc.manuscriptcentral.com/gnte 77.38 Nondestructive Testing and Evaluation Sample 17 11.30 88.70 17.83 82.17 Sample 18 10.68 89.32 14.23 85.77 Sample 19 21.05 78.95 14.39 85.61 Sample 20 15.27 84.73 21.63 78.37 Average 14.98 85.02 22.89 77.11 r Fo er Pe w vie Re ly On 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 58 of 58 URL: http://mc.manuscriptcentral.com/gnte ...Page of 58 SVRNA system – a new tool for automatic classification of microstructure based on backpropagation artificial neural network V H C de Albuquerque 1*, P C Cortez2, A R de Alexandria2,... quantify microstructures of nodular cast iron, malleable cast iron, and gray cast iron, in images The compared computational systems are the SVRNA system (Computational Vision System based on an... and automatically the microstructures of the materials in analysis, what has not always occurred ev with the Image Pro-Plus Thus, we could conclude that our SVRNA system is a valid and ie adequate