Simplified hog computing

Hog has been using it for a long time. Today I tried to write

Most of the functions used are MATLAB toolbox functions.

Where is the gradient histogram difference not written?

The code shows a method for finally displaying a vector graph.

If the program compilation is incorrect, please correct it.

The first is the main program:

clc;clear;image = imread('crop001019.png');figure;imshow(image);image = double(image);image = imresize(image,[128 64]);block = zeros(16,16);cell = 8;block = 16;for i = 0:1:14    for j = 0:1:6        for n = 1:1:16            for k = 1:1:16                block(n,k) = image(i*8+n,j*8+k);            end        end        feature = hist(block);        for e = 1:1:2            for t = 1:1:2                result(i*2+e,j*2+t,:) = feature(e,t,:);            end        end    endendw = result;bs = 8;bim1 = zeros(bs, bs);bim1(:,round(bs/2):round(bs/2)+1) = 1;bim = zeros([size(bim1) 9]);bim(:,:,1) = bim1;for i = 2:9,  bim(:,:,i) = imrotate(bim1, -(i-1)*20, 'crop');ends = size(w);    w(w < 0) = 0;    im = zeros(bs*s(1), bs*s(2));for i = 1:s(1),  iis = (i-1)*bs+1:i*bs;  for j = 1:s(2),    jjs = (j-1)*bs+1:j*bs; %     k = find(w(i,j,:)==max(w(i,j,:)));%     if length(k)>1%         k = k(1);%     end    for k=1:1:9        im(iis,jjs) = im(iis,jjs) + bim(:,:,k) * w(i,j,k);    end  endendscale = max(w(:));im = uint8(im * (255/scale));% im = uint8(im);figure;imshow(im);

The hist is used to calculate the gradient histogram of each block.

Returns a 2*2*9 matrix.

The Code is as follows:

function feature = hist(block)block = block;feature = zeros(2,2,9);for i = 0:1:1    for j = 0:1:1        for k = 1:1:8            for n = 1:1:8                cell(k,n) = block(i*8+k,j*8+n);            end        end        hx = [-1,0,1];        hy = -hx';        Fx = imfilter(double(cell),hx);        Fy = imfilter(double(cell),hy);        Q = atan2(Fy,Fx);        Q = abs(Q);        Q = Q.*180./pi;        Q = round(Q);        for u = 1:1:64            if Q(u)>=0&Q(u)<20 feature(i+1,6) = feature(i+1,6) + 1; end            if Q(u)>=20&Q(u)<40 feature(i+1,5) = feature(i+1,5) + 1; end            if Q(u)>=40&Q(u)<60 feature(i+1,4) = feature(i+1,4) + 1; end            if Q(u)>=60&Q(u)<80 feature(i+1,3) = feature(i+1,3) + 1; end            if Q(u)>=80&Q(u)<100 feature(i+1,2) = feature(i+1,2) + 1; end            if Q(u)>=100&Q(u)<120 feature(i+1,1) = feature(i+1,1) + 1; end            if Q(u)>=120&Q(u)<140 feature(i+1,9) = feature(i+1,9) + 1; end            if Q(u)>=140&Q(u)<160 feature(i+1,8) = feature(i+1,8) + 1; end            if Q(u)>=160&Q(u)<=180 feature(i+1,7) = feature(i+1,7) + 1; end        end    endendfeature = feature./64;

There is something that can be optimized or wrong, please point it out.