Logbook.

Combined result of Shi-Tomasi and S plane

 * Final Image

Hough Line Transform

 * Input Image


 * Houghlines4.jpg


 * Houghlines5.jpg

HSV split planes

 * Input Image


 * H plane


 * S plane


 * V plane


 * Thresholding S plane

Conclusion: The results for this image is satisfactory. All the other output images can be viewed here: https://drive.google.com/open?id=0B18XZ6KmjnivQnFxb1hWdjBxUHM

Removing Green

 * Input Image


 * Output Image

Conclusion: The range in this one is quite limited. So max. green component isn't removed.

Converting to HSV image

 * Input Image


 * Output Image

Observation: If we are to remove the green color entirely, it would affect the image. As observed, some of the cars are shown in green light, so by removing green that information can be lost. This is just converting the image to HSV so as to observe the content of green in image.

Shi-Tomasi corner detection

 * Baltimore11.PNG


 * Baltimore211.PNG


 * Denver1.PNG


 * Detroit11.PNG


 * Detroit211.PNG


 * Detroit311.PNG


 * NJ1.PNG


 * NY11.PNG


 * NY211.PNG


 * NY311.PNG

Observation: Maximum no. of corners are detected in the parking lots. This can be used as one of the process to detect the presence of parking lot.

Using otsu's thresholding and addition with original image

 * Input Image


 * Output Image

Conclusion: Better results than previous examples.

Otsu's Thresholding with Laplacian Filter

 * Input Image


 * Output Image

Conclusion: As image is very unclear and it is hard to detect any object in this, we cannot use this to detect cars.

Basic Blob Detection

 * Ideal Image


 * Image with keypoints

Trouble detecting keypoints and displaying them.

Detecting cross & zeros in tic-tac-toe

 * Input Image


 * Output Image

Detecting shapes in hand drawn image

 * Input Image


 * Output Image

Detecting shapes in paint image

 * Paint Image


 * Gray Image


 * Threshold Image


 * Canny Edge Image


 * Output Image

Output= 4 square 3 triangle 5 pentagon Total no.= 3

Drawing Contours

 * Input Image


 * Output Image

Canny Edge Detection

 * Input Image


 * Output Image