First I tried color segmentation but then the resulting binary image is not good to be processed. The color of the cylinder has almost the same color as the light reflection in the wood. (e.g. see the lower right portion of the image)
Next I tried grayscale thresholding. This one produced a more segmented ROI than the color segmentation so I used this method for the preliminary processing. After thresholding I applied an opening operator using a 5x15 structuring element. The resulting images are shown below.
From the morphed images, I then calculated the center of the ROI and noted its pixel position. The horizontal-pixel location of the center of the ROI represents the displacement or distance traveled of the object. Each frame or image can be translated into time using the fps of the video camera.
The inclined plane has a length of about 480mm which is 240 pixels long in the image. So the distance to pixel conversion is 1px = 2mm. The camera has an fps of 24.5 so 1 frame (or 1 image) is equal to 1/24.5 sec.
I computed the acceleration of the object using the series of processed images using the formula change in velocity per change in time (dv/dt) where velocity is change in distance per change in time(dx/dt). The computed value of acceleration, using the pixel to mm and frame to sec conversion above, is 0.520m/sec. From physics books, we can see that the acceleration of a hollow cylinder rolling in an inclined plane is simple given by a = 0.5*g*sin(theta) where theta is the angle of inclination. In our setup, the angle is approximately 6 degrees so the acceleration is 0.512m/sec. The result from video processing gives an error of 1.6%.
For this activity, I give myself a grade of 10 since I was able to do it properly.
Next I tried grayscale thresholding. This one produced a more segmented ROI than the color segmentation so I used this method for the preliminary processing. After thresholding I applied an opening operator using a 5x15 structuring element. The resulting images are shown below.
From the morphed images, I then calculated the center of the ROI and noted its pixel position. The horizontal-pixel location of the center of the ROI represents the displacement or distance traveled of the object. Each frame or image can be translated into time using the fps of the video camera.The inclined plane has a length of about 480mm which is 240 pixels long in the image. So the distance to pixel conversion is 1px = 2mm. The camera has an fps of 24.5 so 1 frame (or 1 image) is equal to 1/24.5 sec.
I computed the acceleration of the object using the series of processed images using the formula change in velocity per change in time (dv/dt) where velocity is change in distance per change in time(dx/dt). The computed value of acceleration, using the pixel to mm and frame to sec conversion above, is 0.520m/sec. From physics books, we can see that the acceleration of a hollow cylinder rolling in an inclined plane is simple given by a = 0.5*g*sin(theta) where theta is the angle of inclination. In our setup, the angle is approximately 6 degrees so the acceleration is 0.512m/sec. The result from video processing gives an error of 1.6%.
For this activity, I give myself a grade of 10 since I was able to do it properly.
No comments:
Post a Comment