Image Processing
- by Matlab
(Click the picture to see the bigger one)
Thinning
Made from April 4 to April 11, 2013
Thinning is a way to find the bone of the image. The image should be black and white. It's a type of topological skeleton, but computed using mathematical morphology operators.
This is a group project in Image Process Class in a 5-people team. The thing I do is writing the code and test it with several picture.
(Click the picture to see the bigger one)
Improved direction estimation for Di Zenzo's multichannel image gradient operator
Made from November 6 to November 11, 2012
Gradient has been widely used in image processing and computer vision, such as edge detection, image segmentation, corner detection, image fusion, image recognition, face detection, and object tracking. The gradient associated with an image pixel, and usually defined as a 2-D column vector, in which the vectorial angle denotes the direction of the largest growth of the image function. For grayscale images, there’s numerous gradient estimators have been developed. For multichannel (multidimensional) images, this hasn’t received enough attention. For a general point of view in multidimensional gradient estimators, these can be divided into three major categories:
1. A single estimate of the orientation and strength of an edge at a point
2. ( based on grayscale ) The gradient vectors are calculated for individual channels and then combine them to produce the final gradient vectors
3. Find the maximum changes of image vectors
Among these multidimensional gradient estimators mentioned above, perhaps the most classical and widely-used one is Di Zenzo’s multichannel gradient operator. But Di Zenzo didn’t solve the problem of indetermination of the gradient direction. It would results in errors, so we solve this problem thoroughly and at the same time analyze the gradient angle ranges in various cases depend on the same name of the paper written by Lianghai Jin, Hong Liu, Xiangyang Xu, and Enmin Song.
We use Canny edge detection to find the edge. First we use Gaussian filter to remove noise and find back the edge in the image. Then we use improved version of Di Zenzo's gradient operator to find maginitude and direction. And then to use direction to find the edge pixel and high light them.
This is a group project in Image Process Class in a 5-people team. The thing I do is writing the code and test it with several picture.
change the picture into grey (left) and do the Contrast Enhancement (right)
change the picture into grey (left) and do the Contrast Enhancement (right)
change the picture into grey (left) and do the Contrast Enhancement (right)
(Click the picture to see the bigger one)
Contrast Enhancement
Made from November 6 to November 11, 2012
Using Histogram Equalization in Partially Overlap to improve the quality for grey image.
This is a group project in Image Process Class in a 5-people team. The thing I do is writing the code and test it with several picture.
Windyga Fast Impulsive Noise Removal
Made from October 18 to October 25, 2012
This is using the concept of Peak-Valley, which caculate with peak-cutting and valley-filling.
This is a group project in Image Process Class in a 5-people team. The thing I do is writing the code and test it with several picture.