Created Project_Applying_Optical_Flow_to_Determine_the_direction_of_traffic.m file

Computer Vision Specialization/Project_Applying_Optical_Flow_to_Determine_the_direction_of_traffic.m

@@ -0,0 +1,75 @@
+% Paste your solution code from "Project: Applying Optical Flow to Detect Moving Objects" here:
+frame1 = imread("Rt9Frame1.png");
+frame2 = imread("Rt9Frame2.png");
+montage({frame2,frame1})
+%%
+%Calculating the optical flow vector between frame1 & frame2
+
+myOpticalFlow = opticalFlowFarneback;
+estimateFlow(myOpticalFlow,im2gray(frame1));
+flow = estimateFlow(myOpticalFlow,im2gray(frame2)); 
+% myOpticalFlow stores the previous frame
+%%
+%Just to see
+
+imshow(frame2)
+hold on
+plot(flow,"DecimationFactor",[15 15],"ScaleFactor",7)
+hold off
+%%
+%Creating Mask with optical flow magnitude 1+
+
+vm = flow.Magnitude;
+maskThreshold = 1.0;
+mask = (vm(:,:)>maskThreshold);
+%image processing 
+
+se = strel("disk",20,0);
+mask = bwareafilt(mask, [500, inf]);
+mask = imclose(mask, se);
+
+
+% Uncomment below to view your optical flow vectors on your masked image
+maskedFrame = frame2; maskedFrame(repmat(~mask, [1 1 3])) = 0;
+imshow(frame2)
+hold on
+plot(flow,"DecimationFactor",[15 15],"ScaleFactor",7)
+hold off
+figure
+imshow(maskedFrame)
+
+
+% Write new code to count the number of cars moving in each direction and their x-direction velocities
+% Calculate x-direction velocities
+velocities = flow.Vx;
+
+% Calculate average Vx for each object in the mask
+regions = regionprops(mask, 'PixelIdxList');
+avgVxPerObject = zeros(1, numel(regions));
+
+for i = 1:numel(regions)
+    pixelIdxList = regions(i).PixelIdxList;
+    avgVx = mean(velocities(pixelIdxList));
+    avgVxPerObject(i) = avgVx;
+end
+
+% Initialize variables
+leftVx = [];
+rightVx = [];
+numCarsLeft = 0;
+numCarsRight = 0;
+
+% Identify cars moving left or right based on average Vx
+for i = 1:numel(regions)
+    avgVx = avgVxPerObject(i);
+
+    if avgVx < 0 && abs(avgVx) > 3
+        leftVx(end + 1) = avgVx;
+        numCarsLeft = numCarsLeft + 1;
+    elseif avgVx > 0 && avgVx > 3
+        rightVx(end + 1) = avgVx;
+        numCarsRight = numCarsRight + 1;
+    end
+end
+
+%added by pythoner-code(Himanshu).