Colony

%Function which creates a heat map of wall location probabilities.

%Input comes from 'filename' where each line contains 8 space-separated values

%Each line contains: x, y, theta, IR1, IR2, IR3, IR4, IR5 in that order

%Example line of input file:

%17 23 .45 -1 42 -1 -1 -1

%x and y are centimeters, theta is in radians

%IR values are the values seen by the robot, where -1 means that nothing is seen

%Output: probability density map.

%All output measurements are in centimeters

%TODO

%find conversion factor between IR sensor values and centimeters (currently 1 is being used)

function map(filename)

%Vectors from center of bot to IR sensors

Bot_IR1 = [4.5,3.5];

Bot_IR2 = [7.5,0];

Bot_IR3 = [4.5,-3.5];

Bot_IR4 = [-4,-3.5];

Bot_IR5 = [-4,3.5];

Bot_IR = [Bot_IR1; Bot_IR2; Bot_IR3; Bot_IR4; Bot_IR5];

%Unit vectors with the same angle that the IR sensors are tilted at

Angle_IR1 = [.78,.63];

Angle_IR2 = [1,0];

Angle_IR3 = [.78,-.63];

Angle_IR4 = [0,-1];

Angle_IR5 = [0,1];

Angle_IR = [Angle_IR1; Angle_IR2; Angle_IR3; Angle_IR4; Angle_IR5];

%For now, read input in from a file

Data = dlmread(filename, ' ');

points = [];

%Map area:

[X1,X2] = meshgrid(-100:5:100,-100:5:100);

X = [X1(:),X2(:)];

prob_map = zeros(size(X1,1),size(X2,2));

%Set position and range error constant for testing FIXME!

range_error = 250;

pos_error = 500;

%loop through each line of file

i = 1; ir = 1;

%for i=1:size(Data, 1)

    row = Data(i, :);

    x = row(1);

    y = row(2);

    theta = row(3);

	%look at each IR sensor reading

    %for ir=1:5

        %if row(ir+3) ~= -1

            %Find vector of object relative to center of bot

            distance = row(ir+3) * 1; %TODO!!!!!!!!! replace 1 with a conversion factor to cm

            ObjectVector = distance * Angle_IR(ir, :);

            ObjectVector = ObjectVector + Bot_IR(ir, :);

            %rotate by theta

            objR = sqrt(ObjectVector(1)^2 + ObjectVector(2)^2);

            objTheta = atan2(ObjectVector(2), ObjectVector(1));

            objTheta = objTheta + theta;

            %add absolute position and shift into image

            objX = (objR * cos(objTheta)) + x + 1500;

            objY = (objR * sin(objTheta)) + y + 1500;

            ObjectVector = [objX, objY];

			%Calculate distribution for this point:

			%Range variance in y direction, rotated by the object angle.

			rotation = [cos(objTheta),sin(objTheta); -sin(objTheta),cos(objTheta)];

			range_error_cov = inv(rotation)*[0,0;0,range_error]*rotation; 

			%Position variance in the robot direction 

			%(TODO introduce error in position estimation perpendicular to the robot)

			rotation = [cos(theta),sin(theta); -sin(theta),cos(theta)];

			position_error_cov = inv(rotation)*[0,0;0,pos_error]*rotation;

			covar = position_error_cov + range_error_cov;

			dist = reshape(mvnpdf(X,[x,y],covar), size(X1,1),size(X2,2))

			surf(X1,X2,dist);

			colormap default;

			return;

			%Using the values of the pdf as an approximation for the prob. density map

			%(needs to change) update as:

			prob_map = prob_map.*(1-dist) + dist;	

        %end

    %end

%end

%graph the probability map.

surf(X1,X2,prob_map);