Since the images are the same size, this can be done with a simple blend, without the need for any compositing.
% reading in the two images
Protractor = importdata('STS_Protractor2.jpg');
Torso = importdata('STS_Torso2.jpg'); % top of midline is at (349,304) (349.5,316)
RotationPoint = [510,745]; % This is the point the Torso image will be rotated around (center of hip)
Protractor_0 = [510,135]; % location of the 0 value on the protractor
LineLength = 500; % Length of the line centered on the torso and extending past the head
Angle = 25;
timeperiod = 'Pre00';
% rotate trunk image by max trunk angle from model output data, then format image so it can be placed on top of protractor image
TrunkAngle_Rotation = rotateAround(Torso,RotationPoint(2),RotationPoint(1),-Angle, 'crop'); % (image,y-coord,x-cord,angle,'variable') - Allows for rotation of torso image
Clear_Background = ~rotateAround(true(size(Torso)),RotationPoint(2),RotationPoint(1),-Angle); % When image is rotated, this changes the background to white rather than leaving black space.
TrunkAngle_Rotation(Clear_Background) = 255; % Redefines the variable to include the new white background
% blend the two images using a "darken" blend
outpict = min(Protractor,TrunkAngle_Rotation);
% Plotting the output image
imshow(outpict,'border','tight');
Note that there's nothing preventing the hand and torso from obscuring the thigh. That can be addressed.
% reading in the two images
Protractor = importdata('STS_Protractor2.jpg');
Torso = importdata('STS_Torso2.jpg'); % top of midline is at (349,304) (349.5,316)
RotationPoint = [510,745]; % This is the point the Torso image will be rotated around (center of hip)
Protractor_0 = [510,135]; % location of the 0 value on the protractor
LineLength = 500; % Length of the line centered on the torso and extending past the head
Angle = 25;
timeperiod = 'Pre00';
% create a mask
mask = imfill(im2gray(Torso)<250,'holes');
mask = bwareafilt(mask,1);
imshow(mask,'border','tight');
% rotate trunk image by max trunk angle from model output data, then format image so it can be placed on top of protractor image
TrunkAngle_Rotation = rotateAround(Torso,RotationPoint(2),RotationPoint(1),-Angle, 'crop'); % (image,y-coord,x-cord,angle,'variable') - Allows for rotation of torso image
Clear_Background = ~rotateAround(true(size(Torso)),RotationPoint(2),RotationPoint(1),-Angle); % When image is rotated, this changes the background to white rather than leaving black space.
TrunkAngle_Rotation(Clear_Background) = 255; % Redefines the variable to include the new white background
mask = rotateAround(mask,RotationPoint(2),RotationPoint(1),-Angle, 'crop'); % rotate the mask
% blank the background where the FG overlaps
Protractor(repmat(mask,[1,1,size(Protractor,3)])) = 255;
% blend the two images using a "darken" blend
outpict = min(Protractor,TrunkAngle_Rotation);
% Plotting the two figures ontop of each other
imshow(outpict,'border','tight');
... although I suppose that's still kind of wrong. I guess the arm and torso could be masked separately, but I don't know how far you want to go with that. The concept of carrying the mask through the transformation should hold.
Using an RGBA workflow would simplify this a bit, but not with standard IPT/MATLAB tools.
