Feature extraction using crossing number (cn) and ridge tracking technique

Feature extraction using crossing number (cn) and ridge tracking technique PROPOSED ALGORITHM: FEATURE EXTRACTION USING CROSSING NUMBER (CN) AND RIDGE TRACKING TECHNIQUE The various steps involved in feature extraction are as given below: 3.2.1 ADAPTIVE BINARIZATION The enhanced greyscale image is converted to a binary image using adaptive binarization [1]. Global thresholding is not used for binarization because of possibilities of non-uniform illumination on the surface of scanner. Thus using adaptive binarization with a window size of 91 x 91 (This size was finalised after a number of trial and errors). The algorithm can be outlined as follows: Algorithm: Adaptive binarization Input: Enhanced greyscale image e(x,y). Output: Binarized image bin(x,y). For each pixel (i) of e(x,y) Compute local mean (ml) in the 91 x 91 neighborhood of the pixel. If ml > e(xi,yi) then, bin(xi,yi) = white. Else bin(xi,yi)= black. End For. - 3.2.2. THINNING The binarised image is skeletonised using medial axis transformation (MAT)[1] to obtain a single pixel thin ridge structure. The thinning algorithm can be outlined as follows: Assumptions: Region points are assumed to have value 1(white) and background points to have value 0(black). Notations: 1. The 8 neighbour notation of a centre pixel p1 is as shown. p9 p2 p3 p8 p1 p4 p7 p6 p5 2. n (p1) is the number of non zero neighbours of p1. I.e. n (p1) = p2 + p3 + †¦. + p9. 3. t (p1) is the number of 0-1 transitions in the ordered sequence p2, p3,†¦p9,p2. Algorithm : Thinning Input: Binarized image bin(x,y). Output: One pixel thinned image th(x,y). Steps: 1. W.r.t the neighborhood notation a pixel p1 in bin(x,y). is flagged for deletion if the following conditions are satisfied; 2 ≠¤ n(p1) ≠¤ 6 . t(p1)=1. p2 V p4 V p6 = 0 p4 V p6 V p8 = 0 2. Delete all the flagged pixels from bin(x,y). 3. W.r.t the neighborhood notation a pixel p1 in bin(x,y) is flagged for deletion if the following conditions are satisfied; 2 ≠¤ n(p1) ≠¤ 6 . t(p1)=1. p2 V p4 V p8 = 0 p2 V p6 V p8 = 0 4. Delete all the flagged pixel from bin(x,y). 5. Go to step 1 if bin(x, y) is not same as the previous bin(x, y) (indicating that single pixel thickness is yet not obtained) 6. Assign the image bin(x, y) obtained from step 4. to th(x, y). Thus one iteration of the thinning algorithm consists of applying step 1 to flag border points for deletion deleting the flagged points; applying step 3 to flag the remaining border points for deletion; and deleting the flagged points. The basic procedure is applied iteratively until no further points are deleted, at which time the algorithm terminates, yielding the skeleton of the region. 3.2.3 ESTIMATING SPATIAL CO-ORDINATES DIRECTION OF MINUTIAE POINTS. Minutiae representation is by far, the most widely used method of fingerprint representation. Minutia or small details mark the regions of local discontinuity within a fingerprint image. These are locations where the the ridge comes to an end(type: ridge ending) or branches into two (type: bifurcation). Other forms of the minutiae includes a very short ridge (type: ridge dot), or a closed loop (type: enclosure). The different types of minutiae are illustrated Figure 1. There are more than 18 different types of minutiae [2] among which ridge bifurcations and endings are the most widely used. Other minutiae type may simply be expressed as multiple ridge endings of bifurcations. For instance, a ridge dot may be represented by two opposing ridge endings placed at either extremities. Even this simplification is redundant since many matching algorithms do not even distinguish between ridge ending and bifurcations since their types can get flipped. The template simply consists of a list of minutiae location and their orientations. The feature extractor takes as input a gray scale image I(x,y) and produces a unordered set of tuples- M = {m1,m2,m3mN}. Each tuple mi corresponds to a single minutia and represents its properties. The properties extracted by most algorithms include its position and orientation. Thus, each tuple mi is usually represented as a triplet {xi, yi, ÃŽ ¸i}. The crossing number (CN) method is used to perform extraction of the spatial coordinates of the minutiae points. This method extracts the bifurcations from the skeleton image by examining the local neighborhood of each ridge pixel using a 33 window. The CN of a ridge pixel ‘p is given as follows CN=0.5 { i=18pi-pi+1 } p(9) =p(1) . For a pixel ‘p if CN= 3 it is a bifurcation point. For each extracted minutia along with its x and y coordinates the orientation of the associated ridge segment is also recorded. The minutia direction is found out using a ridge tracking technique. With reference to figure 3.3 once the x and y coordinates of the bifurcation point are known, we can track the three directions from that point. Each direction is tracked upto 10 pixel length. Once tracked we construct a triangle from these three points. The midpoint of the smallest side of the triangle is then connected to the bifurcation point and the angle of the resulting line segment is found which is the minutia direction. Assumptions: Ridges are assumed to have value 0 (black) and background points to have value 1(white). Notations: The 8 neighbor notation of a center pixel p1 is as previously shown. The algorithm for extracting the minutiae using the crossing number technique can be outlined as follows: Algorithm: Crossing number Input: Thinned image th(x,y). Output: Image with (x,y) coordinates and orientation thita of each minutia. Steps: 1. For every pixel p in th(x,y) compute the crossing number (CN) ; CN=0.5 { i=18pi-pi+1 } p(9) =p(1) . 2. If CN= 3, the pixel p is declared as a bifurcation point and its x and y coordinates, i.e. p.x and p.y are recorded. 3. The orientation at the bifurcation points p.ÃŽ ¸ is calculated using tracking algorithm. Fingerprint matching Process:- Each minutiae may be described by a number of attributes such as its position (x,y), its orientation ÃŽ ¸, its quality etc. However, most algorithms consider only its position and orientation information. Given a pair of fingerprints and their corresponding minutiae features to be matched, features may be represented as an unordered set given by I1 = {m1,m2.mM} where mi = (xi, yi, ÃŽ ¸i) I2 = {m1,m2.mN} where mi = (xi, yi , ÃŽ ¸i ) Here the objective is to find a point mj in I2 that exclusively corresponds to each point mi in I1. Usually points in I2 is related to points in I1 through a geometric transformation T( ). Therefore, the technique used by most minutiae matching algorithms is to recover the transformation function T( ) that maps the two point sets . The resulting point set I2 is given by: I2 = T(I1) = {m†1,m† 2,m† 3.m†M} m†1 = T(m1) m† N = T(mN) The minutiae pair mi and m†j are considered to be a match only if (xi-xj)2+(yi-yj)2≠¤r0 min( |ÃŽ ¸i − ÃŽ ¸Ã¢â‚¬  j | , 360 − |ÃŽ ¸i − ÃŽ ¸Ã¢â‚¬ j | ) Here r0 and ÃŽ ¸0 denote the tolerance window. The matcher can make on of the following assumptions on the nature of the transformation T Rigid Transformation: Here it is assumed that one point set is rotated and shifted version of the other. Affine Transformation: Affine transformations are generalization of Euclidean transform. Shape and angle are not preserved during transformation. Non-linear Transformation: Here the transformation may be due to any arbitrary and complex transformation function T(x,y). The problem of matching minutiae can be treated as an instance of generalized point pattern matching problem. In its most general form, point pattern matching consists of matching two unordered set of points of possibly different cardinalities and each point. It is assumed that the two pointsets are related by some geometrical relationship. In most situations, some of the point correspondences are already known (e.g. control points in an image registration problem [5,4,6,7])andthe problem reduces to finding the most optimal geometrical transformation that relates these two sets. However, in fingerprints, the point correspondences themselves are unknown and therefore the points have to be matched with no prior assumption making it a very challenging combinatorial problem. There have been several prior approaches where general point pattern techniques havebeen applied. Some of these have been discussed here. Ranade and Rosenfield [8] proposed an iterative approach for obtaining point correspondences. In this approach, for each point pair mi, mj they assign pij , the likelihood of the point correspondence and c(i, j, h, k), a cost function that captures the correspondence of other pairs(mh,m_k) as a result of matching mi with mj. In each iteration pij is incremented if it increases the compatibility of other points and is decremented if it does not. At the point of convergence, each point mi is assigned to the point argmaxk(pik). While this is a fairly accurate approach and is robust to non-linearities, the iterative nature of the algorithm makes it unsuitable for most applications. The hough transform [9] approach or the transformation clustering approach reduces the problemof point pattern matching to detecting the most probable transformation in a transformation search space. Ratha et al [10] proposed a fingerprint matching algorithm based on this approach. In this technique, the search space consists of all the possible parameter under the assumed distortionmodel. For instance, if we assume a rigid transformation, then the search space consists of all possible combinations of all translations (Δx,Δy) , scales s and rotations and ÃŽ ¸. However, to avoid computation complexity the search space is usually discretized into small cells. Therefore the possible transformations form a finite set with Δx ÃŽ µ {Δ1x,Δ2x . . .ΔIx} Δy ÃŽ µ {Δ1y,Δ2y . . .ΔJy} ÃŽ ¸ ÃŽ µ {ÃŽ ¸1, ÃŽ ¸2 . . . ÃŽ ¸K} s ÃŽ µ {s1, s2 . . . sL} A four dimensional accumulator of size (I Ãâ€"J Ãâ€"K Ãâ€"L) is maintained. Each cell A(i, j, k, l) indicatesthe likelihood of the transformation parameters (Δix,Δjy, ÃŽ ¸k, sl). To determine the optimal transformation, every possible transformation is tried on each pair of points. The algorithm used is summarized below for each point mi in fingerprint T . for each point m_j in fingerprint I for each ÃŽ ¸k ÃŽ µ {ÃŽ ¸1, ÃŽ ¸2 . . . ÃŽ ¸K} for each sl ÃŽ µ {s1, s2 . . . sL} compute the translations Δx,Δy Explicit alignment: An illustration of the relative alignment using ridges associated with minutiae mi and mj ∆x∆y=∆xi∆yi-s1cosÃŽ ¸k -sinÃŽ ¸ksinÃŽ ¸k cosÃŽ ¸kxjyj †¦Ã¢â‚¬ ¦Ã¢â‚¬ ¦Ã¢â‚¬ ¦Ã¢â‚¬ ¦Ã¢â‚¬ ¦Ã¢â‚¬ ¦(1) d Let (Δix,Δjy) be the quantized versions of (Δx,Δy) respectively. e If T{mi} matches with m_j increase the evidence for the cell A[Δix,Δjy, ÃŽ ¸k, sl] A[Δix,Δjy, ÃŽ ¸k, sl] = A[Δix,Δjy, ÃŽ ¸k, sl]+1 3.The optimal transformation parameters are obtained using (Δ*x,Δ*y, ÃŽ ¸*, s*) = argmax(i,j,k,l) A[Δix,Δjy, ÃŽ ¸k, sl] References: Gonzalez, Woods, and Eddins. Digital Image Processing using matlab. Prentice Hall, 2004. D. Maltoni, D. Maio, A.K. Jain, S. Prabhakar, Handbook of Fingerprint Recognition, Springer, 2003, ISBN 0-387-95431-7. R.Thai, Fingerprint image enhancement and feature extraction. Australia. Anil Jain, Salil Prabhakar, Lin Hong, and Sharath Pankanti. Filterbank-based fingerprint matching. In Transactions on Image Processing, volume 9, pages 846-859, May 2000. Anil Jain, Arun Ross, and Salil Prabhakar. Fingerprint matching using minutiae texture features.In International Conference on Image Processing, pages 282-285, october 2001. L. Hong, Y. Wang, and A. K. Jain. Fingerprint image enhancement: Algorithm and performanceevaluation. Transactions on PAMI, 21(4):777-789, August 1998. L. Brown. A survey of image registration techniques. ACM Computing Surveys, 1992. A. Ranade and A. Rosenfeld. Point pattern matching by relaxation. Pattern Recognition, 12(2):269-275, 1993. R. O. Duda and P. E. Hart. Use of the hough transformation to detect lines and curves in pictures. Communications of the ACM, 15(1), 1972. N. K. Ratha, K. Karu, S. Chen, and A. K. Jain. A real-time matching system for large fingerprint databases. Transactions on Pattern Analysis and Machine Intelligence, 18(8):799-813, 1996.

Outline for the Good Earth Essay

One Man’s survival and triumph over the land and nature leads to a prosperous life. Thesis: Man’s triumph over the land and nature rewards with wealth and profit and respect from other. II. Introduction- How Wang Lang is connected to the earth and his strong relationship with it and how his good work ethics and moral judgments guide him on becoming one with his land. How Wang Lung tries to establish a connection with the land, the rewards and wealth from having a strong connection with the earth, and the respect from other while leading to a prosperous life. A. Establishing a connection- How Wang Lung attempts to have a strong connection with the earth. 1. Wang Lung starts connecting with the land a. Farms through own physical labor at first but O-Lan help his after they are together. b. He maintains his farm constantly through the changing seasons. 2. Wang Lung respect for nature guides him through his future success, How the nourishing power of the land comforts Wang Lung. b. The Earth producing for Wang Lung for his hard work and dedication B. Disasters for the land- How the nature damages Wang Lung connection with the Earth 1. Wang Lung and his family trying to survive against the elements of nature a. Wang Lung tries to survive from the famine that has struck the village because of the drought and is forced to move away from his land for a while. b. The flood affects Wang Lungs crops but because of his success from farming he is unaffected but becomes severed from his connection with the Earth. Wang Lung is forced to sever his connection with the earth because of nature a. When the famine struck Wang Lung is forced to move away from his land severing his connection and losing his strength to stay upon the land. b. Almost coming to the decision upon selling his daughter to return to the land corrupts Wang Lung moral judgments. C. Triumph over nature and the land- How Wang Lung connection with the land is restored and the wealth he is rewarded with. 1. Wang Lungs connection with the Earth is stronger than before. When returning from the city with money him able to purchase property and profit from his expanding land he has gained because of the Earth providing him with more resources. b. Through hard work he has become more profitable and wealthier than before and is able to provide for his family. 2. Wang Lung leading a prosperous lifestyle with the current wealth he has gained. a. He is able to become wealthier and afford many lavished items in his household and lead a good lavished life while he is now old. b. He becomes well respected within his village and is looked upon as one of the great family’s to the villagers.

How to Register to Vote in U.S. Elections

Registering to vote is required in order to cast ballots in elections in all states except North Dakota. Under Articles I and II of the U.S.  Constitution, the manner in which federal and state elections are conducted are determined by the states. Since each state sets its own election procedures and regulations – such as voter identification laws – it is important to contact your state or local elections office to learn your state’s specific election rules. How to Vote With the exception of state-specific rules, the basic steps to voting are the same almost everywhere.Voter registration is required in every state except North Dakota.Every state allows absentee voting.Most states assign voters to vote at specific polling places or voting locations.The U.S. Election Assistance Commission lists federal election dates and deadlines by state. Who Cannot Vote? The right to vote is not universal. Some people, depending on their circumstances and state laws, will not be allowed to vote. Non-citizens, including permanent legal residents (green card holders), are not allowed to vote in any state.Some people who have been convicted of felonies cannot vote. These rules may vary by state.In some states, persons who have been legally declared mentally incapacitated cannot vote. What is Voter Registration? Voter registration is the process used by the government to ensure that everyone who votes in an election is legally eligible to do so, votes in the correct location and only votes once. Registering to vote requires that you give your correct name, current address and other information to the government office that runs elections where you live. It might be a county or state or city office. Why is Registering to Vote Important? When you register to vote, the elections office will look at your address and determine which voting district you will vote in. Voting in the right place is important because who you get to vote for depends on where you live. For example, if you live on one street, you may have one set of candidates for city council; if you live the next block over, you may be in a different council ward and be voting for completely different people. Usually, the people in a voting district (or precinct) all go to vote in the same location. Most voting districts are fairly small, though in rural areas a district can stretch for miles. Whenever you move, you should register or re-register to vote in order to make sure you always vote in the right place. Who Can Register to Vote? To register in any state, you need to be a U.S. citizen, 18 or older by the next election, and a resident of the state. Most, but not all, states have two other rules as well: 1) you cant be a felon (someone who has committed a serious crime), and 2) you cant be mentally incompetent. In a few places, you can vote in local elections even if you are not a U.S. citizen. To check the rules for your state, call your state or local elections office. College Students: College students who live away from their parents or hometown can usually register legally in either place. Where Can You Register to Vote? Since elections are run by states, cities, and counties, the rules on registering to vote are not the same everywhere. But there are some rules that apply everywhere: for example, under the Motor Voter law, motor vehicle offices across the United States must offer voter registration application forms. Other places required the National Voter Registration Act to offer voter registration forms and assistance include: state or local government offices such as public libraries, public schools, offices of city and county clerks (including marriage license bureaus), fishing and hunting license bureaus, government revenue (tax) offices, unemployment compensation offices, and government offices that provide services to persons with disabilities. You can also register to vote by mail. You can call your local elections office, and ask them to send you a voter registration application in the mail. Just fill it out and send it back. Election offices are usually listed in the phone book in the government pages section. It may be listed under elections, the board of elections, supervisor of elections, or city, county or township clerk, registrar or auditor. Especially when elections are coming up, the political parties set up voter registration stations at public places like a shopping mall and college campuses. They may try to get you to register as a member of their political party, but you dont have to do so in order to register. NOTE: Filling out the voter registration form does not mean that you are actually registered to vote. Sometimes application forms get lost, or people dont fill them out correctly, or other mistakes happen. If in a few weeks you have not received a card from the elections office telling you that you are registered, give them a call. If theres a problem, ask them to send you a new registration form, fill it out carefully and mail it back. The Voter Registration card you receive will probably tell you exactly where you should go to vote. Keep your Voter Registration card in a safe place, its important. What Information Will You Have to Provide? While voter registration application forms will vary depending on your state, county or city, they will always ask for your name, address, date of birth and status of U.S. citizenship. You also have to give your drivers license number, if you have one, or the last four digits of your Social Security number. If you dont have either a drivers license or a Social Security number, the state will assign you a voter identification number. These numbers are to help the state keep track of voters. Check the form carefully, including the back, to see the rules for the place where you live. Party Affiliation: Most registration forms will ask you for a choice of political party affiliation. If you wish to do so, you can register as a member of any political party, including Republican, Democrat or any third party, like Green, Libertarian or Reform. You can also choose to register as independent or no party. Be aware that in some states, if you dont select a party affiliation when you register, you will not be allowed to vote in that partys primary elections. Even if you do not select a political party and do not vote in any party primary elections, you will be allowed to vote in the general election for any candidate. When Should You Register? In most states, you need to register at least 30 days before Election Day. In Connecticut, you can register up until 14 days before an election, in Alabama 10 days. Federal law says that you cant be required to register more than 30 days before the election. Details on registration deadlines in each state can be found on the U.S. Election Assistance Commission Web site. Six states have same-day registration - Idaho, Maine, Minnesota, New Hampshire, Wisconsin and Wyoming. You can go to the polling place, register and vote at the same time. You should bring some identification and proof of where you live. In North Dakota, you can vote without registering. Parts of this article are excerpted from the public domain document I Registered, Did You? distributed by the League of Women Voters.