Digital Image Processing - Report Example

Image processing Your name Name of Assignment 20th July, 2014 Abstract The report undertook to carry out edge detection of images, since edges are boundaries connecting similar surfaces which create discontinuities in image intensity among pixels. Thus is an important characteristic of an image that requires analysis using a different frequency from an image. Detection of these characteristics is essential in data compression, image reconstruction and image segmentation as well as in matching. Three methods edge detection, filtering, canny and Roberts were used. It was observed that canny and Roberts misjudged the pixel of the object however they increased the robustness to noise of the image as there is image variation output. Table of Contents Abstract 2 Introduction 4 Methodology 4 Results 5 Image detection and conversion to rgb2gray 5 The filter and rotating filter 45o 6 Rotating filter 90o 9 Rotating filter 135o 10 Comparison with Canny edge detector 12 Comparison with roberts edge detector 14 Conclusion 16 References 17 Appendix 18 Introduction In obtaining an image that is free of noise filtering is necessary. This done by obtaining the new pixel through use of the filter with the image which identifying the signals in a form of a matrix that would be used. In identifying the signals in identifying the signals we obtained two vectors that were necessary to obtain convolution function. After the vectors, we defined the range at which convolution was going to take. In the equation w is taken as an image which needs to be masked using a filtering matrix which would be obtained in the process of calculation. A mask matrix of 3 by 3 will be determined which will be used in the filtering process. The library functions are used to carry out multiplication as well as additions for each mask. Methodology The image obtained is uploaded to the workspace and the image processing begins by using a filter selected. Here the purpose was to use the filtering method, Roberts and canny functions given to create new pixels for use. This functions aims at developing values that would be used to create new pixels. In the Matlab, multiplication will be used to determine the value of the matrices which will be used in filtering the unwanted frequencies. During the development of the code that was used in solving the function, the input was put in small increments producing solutions which were related to each iteration. This means the iteration worked well during the development of the code. Every iteration was approached with its trade-off in case of delays we waited for the result before inputting new codes. in the coding there were no errors thus, testing the code was found to be unnecessary. In case an error was discovered the resulting chord was to be tested using Matlab iterative statements for the first order iteration and the second order iteration. The do statement is used when unknown iterations are required. Then the initialization is then performed to ensure that the code is well tested. The first case checks whether the program detects an error when the input year is invalid and the month is valid. The second test case checks whether the program detects an error when the input year is just valid but the month is less than one. The third test is similar, but it checks to make sure an error is detected if the month given a valid year. The test cases can be partitioned into two equivalence classes. The first three test cases are invalid inputs, while the last two are valid inputs. Thus, the first three test cases should cause the program to generate an error message. If the program does not generate an error message, then we have exposed a bug. The program should produce correct output for the last two test cases. The tester should make sure that the output produced is indeed correct. In general, it is a good idea to classify all test cases as to whether they are valid inputs and the program should produce valid output, or whether they are valid inputs and the program should generate an error message. Results The image that was used is that of a car number plate were acquired to edge detection. Image detection and conversion to rgb2gray >> I = imread('D:\Documents and Settings\Administrator\My Documents\car number plate.JPG'); this will help in importing the image to work space I1 = rgb2gray (I); will change the clour of the image figure subplot(1,2,1), imshow(I) subplot(1,2,2), imshow(I1) - displaying the images This plots the two images that is the original and changed to one figure as shown below; Figure 1: Original and gray image Filter The filter and rotating filter 45o The following filter was selected Filter= Looking at the above filter, one can note that it has middle column unique. The filter was entered into Matlab and rotated at 45o as >> filter=[-1 -1 0 1 1;-1 -1 0 1 1;-1 -1 8 1 1;-1 -1 0 1 1;-1 -1 0 1 1]; filter45=[0 1 1 1 1;-1 0 1 1 1;-1 -1 8 1 1;-1 -1 -1 0 1;-1 -1 -1 -1 0]; This a filter and rotated filter >> g =imfilter(I1,[filter]/2,'replicate'); this filters the image at 0o mask g45 =imfilter(I1,[filter45]/2,'replicate'); this filters the image at 45o mask subplot(1,2,1), imshow(g) subplot(1,2,2), imshow(g45) The two rows of code plots and displays the image on one figure as shown below Figure 2: filtered by filter and filter45 From the two images average pixels making have disappeared as and other edges are not visible. It can be stated that they are almost similar except at 45o there is some edges which are pronounced. This confirms that the two filters are low pass filters as they allowed only low frequency features to pass through while high pass frequency features were rejected. The lines are thinned into 1 pixel. The two filters do allows high frequency features to pass through the image; therefore the image is affected in terms of color, inversion, opacity and saturation. Looking at the original image and the filtered two images side by side, comparison you will note that the arrays of the image changes are the edges not at the centre. It maintains the information of the images as the original. This is designed to reduce the brightness of the image. This kind of filter can be used to improve the quality of an image which has shadows. An image develops shadows when it is taken under too much light. This technology is found in cameras. Rotating filter 90o >> filter90=rot90(filter); totating the filter 90o >> g90 =imfilter(I1,[filter90]/2,'replicate'); filtering the image >> figure >> subplot(1,2,1), imshow(g) subplot(1,2,2), imshow(g90) plotting and displaying the image Tracing along the edge with low will connect the broken edges or restore segments removed by high. For filter90 has an image that is more blurred than filter. This makes it a high pass filter. Images subjected to high pass filters are saturated and the colors are altered. This can be used in reducing the opacity or inverting an image Figure 3: filtered by filter and filter90 When looking at the image that is produced after its filtered there is eyestrain. It is good for viewing images like clouds and the sky. This filter can be used to improve the quality of an image that has been taken in outdoor. However, if care is not taken, a good image can be blurred and become useless if filtered. Rotating filter 135o >> filter135=rot90(filter45); >> g135 =imfilter(I1,[filter135]/2,'replicate'); >> figure >> subplot(1,2,1), imshow(g) >> subplot(1,2,2), imshow(g135) Figure 4: filtered by filter and filter135 Filter135 is high pass filter because the filtered image is blurred and its features are distorted. High pass filters allows high frequency features to pass through and are detected in the image. It also affects the edges of the image. Looking at the contrast between the original image and the filtered image, one will note a difference in the color of the image. There is also a departure in terms of the shadow of the image. The application of this filter is when one wants to reduce the sharpness and the color of an image. It can render a colored image to take black and white element. High pass filters is important because it is used in detecting the edge of the image. It can also be used in computing at derivative of the image that is at what rate the image change does. Most black and grey areas will be visible. In this case, the accuracy/normalization of the image is reduced and non-linear features were visible. Comparison with Canny edge detector The image was read int work space as follows >> I = imread('D:\Documents and Settings\Administrator\My Documents\car number plate.JPG'); This will help in importing the image to work space I1 = rgb2gray (I); will change the clour of the image Figure subplot(1,2,1), imshow(I) subplot(1,2,2), imshow(I1) - displaying the images BW = edge(I1,'canny'); specifies the Canny method >> figure >> imshow(BW) Figure 5: canny The image shows the edges clearly which means they are low pass filters has been used to detect the edges thus reducing the possibilities of having negative features in the image. In any image, edge detection is very important as the users are able to detect the boundaries when using it in making decisions. This is important to surveyors and other professionals. When an image is filtered using high pass filter, some sections of the image are pronounced clearly. Comparison with roberts edge detector All edges of the image are detected; it will take an ordinary eye to differentiate the original image and the filtered image. The vertical and horizontal edges of the image are not altered and are not sharpened as in the case of high pass filter. I = imread('D:\Documents and Settings\Administrator\My Documents\car number plate.JPG'); This will help in importing the image to work space I1 = rgb2gray (I); will change the clour of the image Figure subplot(1,2,1), imshow(I) subplot(1,2,2), imshow(I1) - displaying the images BW = edge(I,'roberts') shows the method used >> figure >> imshow(BW) Figure 6: Roberts The images produced, have a great difference from the original image. The images are blurred and their colors have been altered as images have pixel discontinuities. As it can be noted from the above images, it has its pixels harmonized and it has almost equal features with the original. It is a low pass filter and low pass filters help in producing an image that is clear by averaging or image pixels to create uniformity in the image. This can be seen from the two images above as the section where the forward foot is just about to land has changed from the original to have same pixels as other parts of the picture. Low pass filters are used to produce equipments which are used to make pictures clearer if taken in sunny environments. It removes high frequency features from the image environment. Most cameras have been made using this technology thus their ability to take clear photos in any environment. In some instances, images that are degraded and blurred are subjected to low pass filters to make them clearer. Conclusion Rotating a filter can also avoid the pixels that near to an edge be recognized as an edge pixel, which is usually an important problem when using the filters or functions for edge detection. References Bergenudd, E., 2006, Low Cost Real Time License Plate Recognition for a Vehicle PC’, Master’s Degree Project. Gonzalez RC & Woods, RE, 2008, ‘Digital Image Processing, Instructor's Manual’ New York: Prentice Hall. Gonzalez, RC, Woods, RE and Eddins, SL, 2009, Digital Image Processing Using MatLab’, New York: Prentice Hall. Appendix Read More