Rough-Fuzzy Granularity in the study of optical phenomena Ana Lucia Dai Pra, Lucía Isabel Passoni Faculty of engineering, National University of Mar del Plata, Mar del Plata, Argentina daipra@fi.mdp.edu.ar Abstract a suitable resolution and in stable conditions [Rabal and Braga, 2008]. Granular computing deals with information In ecography images, the speckles are a nuisance that is representation in the form of a number of entities desired to diminish to improve recognition and resolution or information granules. Information granules are made up of a collection of entities, usually of [Damerjiana et al., 2014], [Hiremath et al., 2013] . numeric level, joined due to their similarity, In both cases, laser and ultrasound, a stable speckle pattern functional adjacency, indistinguishability, is achieved when the scatterers don´t move however when coherence or alikeness. The granular computing is the sample has certain activity it is translated to the associated to sets concepts, such as fuzzy sets, scatterers movement, the dynamic of the speckle pattern is rough sets, intervals. used to evaluate the scatterer movements. In this work is considered the application of In this work we propose using a method based on rough- Rough-Fuzzy Granularity to the detection of fuzzy granular computing to detect regions of interest in optical phenomena in registered videos. Basically ecographies and Speckle image stack and also in single these phenomena are dynamic laser speckle and frames, since each of them would allow detecting different ecography videos. types of features. Then, temporal and spatial granularity is . analyzed. . 1 Introduction 2 Methodology In granular computing the information is grouped in entities that fulfill conditions of similarity, being the granules, 2.1 Granular Computing conceptual entities that emerge as a direct consequence of In granular computing the information is grouped in entities the quest for the identification of abstract objects and its that fulfill conditions of similarity, being the granules, processing [Zadeh, 1970], [Pedrycz, 2001], [Yao, 2004]. conceptual entities that emerge as a direct consequence of The fuzzy sets and rough sets are a suitable way to define the quest for the identification of abstract objects and its information granules. processing.. The fuzzy sets and rough sets are a suitable way The identification of information granules is context to define granules. dependent and is expected that achieve two intuitive In the signal processing, the information granules contribute requirements: Justifiable granularity and Semantic to condensing a signal and represent it as a set of temporal meaningfulness. For numeric data, the requirement of granules through an abstraction mechanism that synthesizes Justifiable granularity is quantified by counting the number the information. This representation preserves the granules of data falling within the bounds of the granule, and the identity in spite of some small fluctuations occurring within requirement of semantic meaningfulness is quantified by the the experimental data [Bargiela and Pedrycz, 2003] . This length of the granule [Wang et al., 2015]. type of condensation moves the signal from the numeric The dynamic laser speckle and ecography videos level up to the symbolic processing layer. The granules size exhibit special characteristics. and quantity implies a level of abstraction that is achieved Speckle is a phenomenon that allows to detect activity in (figure 1). several objects through the lighting with laser beams. In spatial granulation the individual pixels of an image are Speckle generate an interference pattern formed by coherent arranged into larger entities and processed as such or density radiation of a medium containing many sub-resolution pixels with determined characteristic are analyzed in small scatterers in move. In order to register the phenomenon, windows. In small windows, the image is built with pixels successive images can be obtained with CCD cameras with computed as the sum of elements quantity belonging to the same fuzzy concept in a slide window of NxN, where the pixel is in the left-upper side. Thus, the image does not lose defining its possible attributes, the subsets X ⊆ U and B⊆ resolution, only it loses an edge of N-1 pixels (figure 2). A, and an equivalence class [x] B which defines a relation in which elements in X are indiscernible from each other by attributes from B, a rough set is defined by the sets {x∈U:[x] B ⊆ X} and {x∈U:[x] B ∩ X ≠ ∅} which are denominated as B-lower approximation and B-upper approximation of X in B respectively. These rough sets are denoted as BX and B X respectively. The objects in BX are certainly members of B and the objects in B X are possibly members in B [Pawlak, 1982], [Pawlak and Skowron, 2007]. If B is a fuzzy attribute, the sets are rough-fuzzy sets [Jensen, 2002] To define the fuzzy-rough regions for each Ck concept, image intensity histogram is analyzed to find an equitable distribution for the number of pixels that will correspond to Figure 1. Temporal Fuzzy Granulation each lower approximation and boundary region (figure 3). To granulate a signal TS(x,y) of length n, corresponding to the (x,y) pixel, successions of equal equivalence classes [i] Ck are considered (upper approximation). A granule ends when an 12 Dark granules equivalence class membership is zero. 9 Medium granules 12 Light granules Figure 2. Spatial Fuzzy Granulation 2.2 Selection of Fuzzy-Rough sets The theory of fuzzy sets, that permits the handling of vagueness and overlapped concepts, makes easy the adequate definition of intensity grains as they are inherent in Figure 3. Rough Parameters selection speckle phenomena. By definition [Zadeh, 1965], [Dubois and Prade, 1980], given a Universal set U of elements ui , a  1 if TS ( x , y ) ( j −1) C = 1 and TS ( x , y ) ( j ) C = 0 Gr( x , y ) ( j , k ) =  k k (1) fuzzy set A∈U is defined by pairs of elements (ui, µA(ui)),  0 in other case where µA(ui) is a real value in [0,1] that represent the with j =1, n and k =1,2,3 membership degree of ui to A. In this case, the U set is given by intensity values I(x,y) ∈ [0,255], and the fuzzy set are The Temporal Rough-Fuzzy Granularity (TRFG) is com- defined by membership functions µ Ck(I(x,y)), with C ∈ puted as the granule quantities Gr in j= n time for k equiva- {dark, medium, light} conceptual sets that define lence classes. Eq (1) and (2) characteristics of the intensity pixels. These fuzzy sets facilitate the interpretation of subjective terms with  3 n  indefinite limits. TRFG( x , y ) =  ∑ ∑ Gr( x , y ) (i , j ) / n (2) A rough set is an approximation of a vague concept by a  k =1 j = 2  pair of precise concepts. Rough sets are based on the fact that an object cannot always be defined in precise form The Rough-Fuzzy spatial granularity (SRFG) is computed as (crisply) inside a category on the basis of the value of its the relative pixels quantities corresponding to equivalence attributes. classes [i] Ck in a window of m*m, where P(x,y) indicate a Formally a rough set is expressed as: pixel. Eq (3) Given an information system S = (U,A), with U the universal set defining all the objects to consider, the A set  3 m m  the inspection of biological, physical and / or chemical SRFG( x , y ) =  ∑ ∑ ∑ [P(x + i , y + j )]Ck  / (m * m ) (3) processes [Todorovich et al., 2013].  k =1 i = 0 j = 0  The computed pixel value will be greater when the window 3.2 Ecography pixels belong to the boundary regions (the pixel belongs to Usually, speckles observed in ultrasound images are only an more than one fuzzy concept). This feature could be artifact, a nuisance that is desired to diminish to improve interpreted as corresponding to the blurring of moving recognition and resolution. In that direction most efforts regions. were directed. To improve the performance of that technique, the instruments are usually provided with filters that smooth slow speckle motions effects, thus avoiding the 3 Experiments dynamics that constitute speckle. 3.1 Speckle In both cases, laser and ultrasound, a stable speckle pattern Speckle is an optical phenomenon that takes place when a is achieved when the scatterers don´t move, therefore the beam of coherent light (laser) illuminates an object with a dynamic of the speckle pattern is used to evaluate the surface that is rough in comparison with the wave length. scatterer movements. Light is scattered in all directions and an interference pattern Activity measurement has shown a noticeable increase in of granular aspect, called ‘speckle pattern’ can be observed research interest in the last few years. Laser speckle patterns on a screen. When the object under study presents some show dynamic behavior when one or more mechanisms act type of activity, such as biological specimens or certain on the observed sample, such as: Doppler Effect, diffusion, physical phenomena, the particles of the surface move and optical polarization activity [Briers, 2007]. the speckle pattern changes over time. This change permits the detection and segmentation of different activity degrees in a diversity of phenomena, allowing us to analyze paint 4 Coments and results drying time, imperceptible bruises in fruits, viability of Figure 4 a) shows a speckle patterns of a corn seed.. Figure seeds, bacteria mobility, endosperm phase proportions, etc. 4 b) shows the identification of vital region of the seed [Rabal and Braga, 2008], [Briers, 2007]. (radicle and embryo). It´s important in the agricultural trade. Stacks of records of successive images are obtained with CCD cameras with a suitable resolution and in stable conditions. The variation of each pixel value over time, considered as a time series, can be analyzed by applying different technologies of signal processing to obtain values that describe its behavior. The set of the values or descriptors obtained in every pixel generates an image whereby it is possible to detect regions with different characteristics. Many of the methods has been studied to analyze dynamic speckle patterns require a high number of images to obtain good results. In some cases, the time of evolution of the analyzed activity is not known a priori and changes inside the time required for the register are lost. Also, records taken beyond the end of the studied events reduce the efficacy of the analysis because of the recording of assumed activity that has already finished, changed or reduced. Non- stationary phenomena cannot be detected. a) b) require the register of images stack where the time history Figure 4. Corn seed of each pixel is analyzed as a time series [Rabal and Braga, 2008]. Less frequent, single frame estimation techniques, such as local spatial contrast measurements [Briers, 2007] have been reported and used in actual applications with the advantages of being able to follow non-stationary processes. The temporal granular computing has been applied to obtain descriptors in dynamic speckle, it provided satisfactory results in the detection of regions with different activity characteristic [Dai Pra et. al, 2009], besides, can perform almost real-time analysis of unquestionable importance in across a dynamic monitoring, optimizing the strategy Figure 5. Painted coin ventilatoria during the general anesthesia. The processing of the video would allow the quantification of the re-aeration Figure 5 shows the result of the processing of a coin with a and of the loss of pulmonary aeration, in order to contribute coat of fresh paint. The zones with relief have more activity to not subjective elements in the interpretation of results. and this allows to visualize the different reliefs of the coin. This experiment is very useful in the study of times of dried of paintings. Figure 6 a) shows a ecography of eye tumor. Figure 6b) is the result of a temporal process, a semi posterior oval region with, seemingly, light dots in the middle that can be estimated could be blood vessels of neo-vascularization (Doppler effect), making very simple the differentiation between both pathologies. Figure 6c) is the result of a spatial process, retina can be seen that cannot be perceived in the other images. Figure 7. Pulmonar ecography References [Al-Hmouz et.al., 2015] Rami Al-Hmouz, Witold Pedrycz, Abdullah Balamash. Description and prediction of time series: A general framework of Granular Computing, Expert Systems with Applications 42: 4830–4839. 2015. [Bargiela and Pedrycz, 2003] Bargiela A, Pedrycz W Granulation of temporal data: a global view on time se- a) ries, in: IEEE Proceedings of the 22nd International Conference of the North American Fuzzy Information Processing Society. 24–26 July, pp. 191–196. 2003. [Briers, 2007] J. D. Briers, Laser speckle contrast imaging for measuring blood flow, Optica Applicata 37 139- 152M. 2007 [Dai Pra et. al, 2009] A. L. Dai Pra, L. I. Passoni, H. Rabal. Evaluation of laser dynamic speckle signals applying granular computing. Signal Process 89 266-274. 2009 [Damerjiana et al., 2014] V. Damerjiana, O. Tan- b) kyevycha, N. Souagb, E. Petita Speckle characterization methods in ultrasound images – A review ,IRBM. Volume 35, Issue 4, September 2014, Pages 202–213 [Dubois and Prade, 1980] Dubois D, H. Prade H. Fuzzy Sets Systems, Academic Press, New York. 1980 [Hiremath et al., 2013] P.S. Hiremath, Prema T. Akkasali- gar and Sharan Badiger. Speckle Noise Reduction in Medical Ultrasound Images in Advancements and Breakthroughs in Ultrasound Imaging", book edited by Gunti Gunarathne, Intech. 2013 c) [Jensen, 2002] R. Jensen, Q. Shen, Fuzzy-rough sets for descriptive dimensionality reduction, Fuzzy Systems, 2002. FUZZ-IEEE'02. Proceedings of the 2002 IEEE In- Figure 6. Eye melanoma ternational Conference on 1: 29-34. 2002. [Pawlak, 1982] Pawlak Z Rough sets. Int J Comput Inf SciI, Figure 7 shows pulmonar ecographies that allow to evaluate 11:341-356. 1982. the efficiency of the maneuver of alveolar recruitment [Pawlak and Skowron, 2007] Pawlak Z, Skowron A Ru- diments of rough sets. Inform Sciences 177: 3–27. 2007. [Pedrycz, 2001] Pedrycz W Granular Computing: An Emerging Paradigm, Phisica- Verlag, Germany, 2001 [Rabal and Braga, 2008] H. J. Rabal, R. A. Braga (eds.): Dynamic Laser Speckle and Applications. CRC Press, 2008. [Todorovich et al., 2013] E. Todorovich, A. L. Dai Pra, L. I. Passoni, M. Vazquez, E. Cozzolino, F. Ferrara, G. Bioul. Real-time speckle image processing. J Real-Time Image Proc. 2013. [Wang et al., 2015] Weina Wang, Witold Pedrycz, Xiaodong Liu. Time series long-term forecasting model based on information granules and fuzzy clustering. En- gineering Applications of Artificial Intelligence. 41: 17– 24. 2015. [Yao, 2004] Yao Y Partition Model of Granular Comput- ing. Transactions on Rough Sets I. Lecture Notes in Computer Sci-ence Volume 3100: 232-253. 2004 [Zadeh, 1965] Zadeh L A Fuzzy sets. Inform. Control 8: 338–353. 1965. [Zadeh, 1970] Zadeh L A Toward a theory of fuzzy infor- mation granu-lation and its centrality in human reason- ing and fuzzy logic, Fuzzy Sets Syst. 90:111–127. 1970