How can I quantify or compare the movements of similar type of complex objects from 2 different videos? (Suppose I saved a particular frames from the videos).

I have 2 separate videos of similar type of objects moving slowly (back and forth or side-by-side). I want to compare and quantify the movements from 2 different videos. The videos are kind of black and white. One idea came to my mind: colormapping the videos/particular frame to indicate movement. Initially, I am trying to detect the edge, but from comparion seems unclear from edge detection. The code that I am using:

% dir = 'C:\Asif\Post processing';

% [fileName, filePath] = uigetfile({'*.*'},'MultiSelect', 'on', '', dir);

%

% if isequal(fileName, 0); errorMsg; return; end; [~, n] = size(fileName);

% if ischar(fileName)==1; n = 1; fileName = cellstr(fileName); end

%

% vr = VideoReader(fullfile(filePath, string(fileName)));

% vid = vr.read();

% vid = vid(:,:,:,1);

% imshow(vid)

ear = imread("ear.jpg");

ear = rgb2gray(ear);

[earEdge,thresh] = edge(ear,"log");

newEdge = edge(ear,"log",thresh/0.45);

imshow(newEdge)

Answer 1

[Aastha]

Hi Md. Asif Hasan,

I understand you are trying to quantify motion of similar objects from two different videos using colour-mapping technique.

You can achieve this by estimating the optical flow of each moving object. By estimating optical flow between video frames, you can measure the velocities of objects in the video and then compare them.

You can estimate the optical flow in MATLAB using “estimateFlow” function that estimates the optical flow of the input video.

Here is the sample code where I compute optical flow for the video "video.mp4":

vidObj = VideoReader('video.mp4','CurrentTime',15); % Create a Video Reader object to read frames from the video ‘video.mo4’. Specify the time-stamp using the CurrentTime name-value argument 

of = opticalFlowHS;  % Specify the optical-flow method you want to use. 

figure;  % create a figure to visualize the motion vectors 

rgbframe = (readFrame(vidObj)); % read a frame from the video  

grayframe = im2gray(rgbframe);  % convert the rgb frame to grayscale 

flow = estimateFlow(of,grayframe); % estimate the optical-flow and store it in the flow variable 

imshow(rgbframe) % visualize the RGB frame 

hold on 

plot(flow,'DecimationFactor',[10 10],'ScaleFactor',60); % plot the flow vectors 

hold off 

title('Frame Motion Visualization') 

 

Refer to the following MathWorks documentation link for more information on

• optical flow

https://www.mathworks.com/discovery/optical-flow.html?s_tid=srchtitle_support_results_1_optical%20Flow

• “estimateFlow” function

https://www.mathworks.com/help/vision/ref/opticalflowhs.estimateflow.html?s_tid=srchtitle_site_search_1_estimateFlow

Hope this helps!