coursework help
in [Matlab]
|
From: Tarun on 5 May 2010 11:26 The question i have in hand with me is as follows Using the image data given in bmp and matrix form in the attached files, you are required to produce a software program implementing the Radix-2 representation method for median filtering. The dimensions of the matrix/image are 256x256. You should consider neighbourhoods of size 3x3, 5x5 and 7x7. You are required to submit: a) The software code in MATLAB; b) A brief report (no longer than three pages), describing the workings of the algorithm, and its effect on noise removal. Comment on the effect of varying neighbourhood sizes; c) The resulting image data in matrix form; d) Display the original and resulting images using e.g., the mat2gray/imshow MATLAB commands (Image Processing Toolbox). and description for the radix 2 filter which is similar to medial filter is RADIX-2 REPRESENTATION ALGORITHM Step 1: Select L (number of bits to represent gray level values), and N (number of pixels in neighbourhood). Step 2: For iteration k=1, select bit index mk=0 and set threshold tk=ceiling(N/2). Step 3: For k=1 to L Do { Calculate members set Π[2mk+1]k. Calculate cardinality a(mk,k). If a(mk,k)≥tk then bit median is μk=1 and set tk+1=tk If a(mk,k)<tk then bit median is μk=0 and set tk+1=tk-a(mk,k) mk+1=2mk+μk } Step 4: Neighbourhood median is given by {μ1, μ2, …, μk} i need to submit this coursework by tommorow evening and i havent even started yet...IM in trouble because my exam are starting next week and i have to study for them....please some one help me with the matlab codes ImageAnalyst <imageanalyst(a)mailinator.com> wrote in message <78649ee1-549e-4f4d-bd3c-8fdc9711a568(a)i9g2000yqi.googlegroups.com>... > On May 5, 10:47 am, Walter Roberson <rober...(a)hushmail.com> wrote: > > ImageAnalyst wrote: > > > I never heard of a radix 2 representation of an image. > > > > "radix 2" would be equivalent to "binary". > > ------------------------------------------------------- > Yes but a binary image is not the same as, say, representing a pixel > with gray level of 169 as 10101001 which is the binary representation > of 169. All images are ultimately binary in that sense because > they're numerical representations in a digital computer. And if you > say > binaryImage = grayImage > 42; > you can't say that binary image is a radix 2 representation of gray > image because it's not - it's derived from grayImage but it is not a > representation of grayImage. So I'm still unclear as to what Tarun > means. I guess we'll see if Tarun returns......
From: Mark Shore on 5 May 2010 11:37 "Tarun " <tarunjacob(a)gmail.com> wrote in message <hrrcr2$6fk$1(a)fred.mathworks.com>... > Please some one can help me with my Image Processing coursework its radix 2 representation of a 256x256 image in matlab ..i have to sumbit the coursework by this friday and i have exams next week...i am not that good in matlab and im having lot of trouble..if some one can help i please mail me on tarunjacob(a)gmail.com and i will send the details....i will be very thankfull...please help... .... > The question i have in hand with me is as follows > Using the image data given in bmp and matrix form in the attached files, you are > required to produce a software program implementing the Radix-2 representation > method for median filtering. The dimensions of the matrix/image are 256x256. You > should consider neighbourhoods of size 3x3, 5x5 and 7x7. You are required to > submit: > a) The software code in MATLAB; > b) A brief report (no longer than three pages), describing the workings of the > algorithm, and its effect on noise removal. Comment on the effect of > varying neighbourhood sizes; > c) The resulting image data in matrix form; > d) Display the original and resulting images using e.g., the mat2gray/imshow > MATLAB commands (Image Processing Toolbox). > > and description for the radix 2 filter which is similar to medial filter is > RADIX-2 REPRESENTATION ALGORITHM > Step 1: Select L (number of bits to represent gray level values), and N (number of pixels in neighbourhood). > Step 2: For iteration k=1, select bit index mk=0 and set threshold tk=ceiling(N/2). > Step 3: For k=1 to L > Do { > Calculate members set Π[2mk+1]k. > Calculate cardinality a(mk,k). > If a(mk,k)≥tk then bit median is μk=1 and set tk+1=tk > If a(mk,k)<tk then bit median is μk=0 and set tk+1=tk-a(mk,k) > mk+1=2mk+μk > } > Step 4: Neighbourhood median is given by {μ1, μ2, …, μk} > i need to submit this coursework by tommorow evening and i havent even started yet...IM in trouble because my exam are starting next week and i have to study for them....please some one help me with the matlab codes > Well, maybe you get a few points for being direct, but as a quick look through CSSM will show, nobody will do your homework and assignments for you (let alone write your exams). On the bright side, failure CAN be a learning experience.
From: Tarun on 5 May 2010 12:46
Hi, My friend and me have been working on a program today and it seems to be stuck with errors....we are currently running out of time...please someone help us with debugging it...we have used a 256 by 256 image and a 3 by 3 mask(window) the pgm code is given below and the question as well...i have also attache the theory for radix 2... a=imread('lena.bmp'); [r c]=size(a); win_size=3; for i=1:r-2, for j=1:c-2, cnt=1; for k=i:i+(win_size-1), for l=j:j+(win_size-1), b=dec2bin(a(k,l),8); for lp=1:8, arr(cnt,lp)=b(lp); end cnt=cnt+1; end end m=0; t=5; for kn=1:8 %steps 1-8 p=(2*m)+1; cn=0; for x=1:9 for y=1:kn xy(x,y)=arr(x,y); end end xy_d=bin2dec(xy); for d=1:9 if (xy_d(d)==p) cn=cn+1; end end if (cn>=t) U(kn)=1; m=(2*m)+1; else if (cn<t) U(kn)=0; m=(2*m)+0; t=t-cn; end end end u(i,j)=U(1)*2^7+U(2)*2^6+U(3)*2^5+U(4)*2^4+U(5)*2^3+U(6)*2^2+U(7)*2^1+U(8)*2^0; end end "Mark Shore" <mshore(a)magmageosciences.ca> wrote in message <hrs3av$k2q$1(a)fred.mathworks.com>... > "Tarun " <tarunjacob(a)gmail.com> wrote in message <hrrcr2$6fk$1(a)fred.mathworks.com>... > > Please some one can help me with my Image Processing coursework its radix 2 representation of a 256x256 image in matlab ..i have to sumbit the coursework by this friday and i have exams next week...i am not that good in matlab and im having lot of trouble..if some one can help i please mail me on tarunjacob(a)gmail.com and i will send the details....i will be very thankfull...please help... > ... > > The question i have in hand with me is as follows > > Using the image data given in bmp and matrix form in the attached files, you are > > required to produce a software program implementing the Radix-2 representation > > method for median filtering. The dimensions of the matrix/image are 256x256. You > > should consider neighbourhoods of size 3x3, 5x5 and 7x7. You are required to > > submit: > > a) The software code in MATLAB; > > b) A brief report (no longer than three pages), describing the workings of the > > algorithm, and its effect on noise removal. Comment on the effect of > > varying neighbourhood sizes; > > c) The resulting image data in matrix form; > > d) Display the original and resulting images using e.g., the mat2gray/imshow > > MATLAB commands (Image Processing Toolbox). > > > > and description for the radix 2 filter which is similar to medial filter is > > RADIX-2 REPRESENTATION ALGORITHM > > Step 1: Select L (number of bits to represent gray level values), and N (number of pixels in neighbourhood). > > Step 2: For iteration k=1, select bit index mk=0 and set threshold tk=ceiling(N/2). > > Step 3: For k=1 to L > > Do { > > Calculate members set Π[2mk+1]k. > > Calculate cardinality a(mk,k). > > If a(mk,k)≥tk then bit median is μk=1 and set tk+1=tk > > If a(mk,k)<tk then bit median is μk=0 and set tk+1=tk-a(mk,k) > > mk+1=2mk+μk > > } > > Step 4: Neighbourhood median is given by {μ1, μ2, …, μk} > > i need to submit this coursework by tommorow evening and i havent even started yet...IM in trouble because my exam are starting next week and i have to study for them....please some one help me with the matlab codes > > > > Well, maybe you get a few points for being direct, but as a quick look through CSSM will show, nobody will do your homework and assignments for you (let alone write your exams). On the bright side, failure CAN be a learning experience. |