LAB 3 Essay Example
Moving average smoothers:
It is a type of spatial domain image filtering used to remove or decrease the noise on an image by averaging the value of each pixel with its neighborhood pixel value.
The aim of this exercise is to see the affect of moving average smoothing filter on the noisy image and what will happen to the quality of the image after using this type of filtering.
Exercise (1: 1)
From the images above (fork images and x-ray chest images), the outcome images of this exercise are original ,noise and smoothed images .The original images of the fork and x-ray chest images have good contrast , clear edges and there are more details in these images. On the other hand, there are some noises it can be seen clearly in the (noise images of the fork and chest x-ray images). So, we can reduce this noise by using the moving average filter on the noisy images to make it smooth and reduce or remove the noise in the images. By applying different numbers of span, the results also were different. So, when the span was 5 to smooth the noisy images, it is clear in both images ( fork and x-ray chest images ) become smoothed and the noise in booth images is decreased while the details of images were vague. However, with the moving average filter, the edges of both images become blurring. Moreover, the result of the images with 5 span is less detail and contrast when compare that with the original images. But it is considered as the better when compare the smoothed images with noise images in both exercises .on the other hand, the result of smoothed images with 5 span is better than the smoothed images when span were 30 and 50 because the details, contrast and the edges of the images have disappeared. Furthermore, when the number of span increased to 30 the smoothed images become blurring and lost more of its details but it can be difficult to recognize the image what it was. However, by increasing the span to 50, the smoothed images lost most of its details , contrast and the image edges become unclear.
The result of this exercise is that the moving average filter is used to make the image smooth by decrease the amount of intensity between the pixels. The main function of the moving average filter is to decrease the image noise. So, if the number of span has increased, the image details will become unclear (blurring).
Exercise 1.2:- ( with edge truncate )
In these 2 images with edge truncate command the result of images are still have blurring( unclear)and the contrast is also decreased when the span value is increased. So, the high span value can effect on the contrast and details of the image and make the edge of the image unclear. Moreover, from the images above, it can be seen clearly that, the noise is decreased when the span value increased but, the smoothing and blurring covered all the details in these images.
(Without edge truncate)
These images above with moving average filter without edge truncate. These two images has different span values. Moreover, the noise is shown on these images and the edges are not clearing (smooth). The result of that is when the filter is small, the edge will be smaller. In the contrary, when the filter is large the edges will become larger. Furthermore, the image centre becomes smoother than the previous images with edge truncate.
The result of this exercise is that: The unsmoothed edges increased when the span value is increased and the image become more obscure and lost most of its details.
Exercise 1.3:- (with edge truncate and span 20)
The images above shows the Original Images, noise images and two images with different span values ( filter size 20 and 35). The result of filtered images is vague and smoothed. Furthermore, the filtered image with 35 span has lost most of its details and the edges are not clear (vague) when compare it by filtered image with 20 span. Moreover, the edges are blurred in both filtered images. On the other hand, the noisy in both filtered images is decreased especially in the image with span 35. Furthermore, the smoothed
images have black and white colours than the noisy images and original. On the other hand, the original image has the best contrast and details.
without edge truncate and span 20-
This figures show the smoothed image by using — without edge truncate- command. The result is clear, when the span value is low (20), the smoothed image has unclear details ( blurred) and cover most area in the smoothed image. Moreover, the edge of smoothed image not effects with blurred. On the other hand, when compare this result with edge truncate command, the blurring and noise cover all the image. Furthermore, the edge of the smoothed image has less noise and blurred. In contrast, When the span value
is high ( 35 ) the edge of smoothed image is large and noise, ( it is clear when you look to the center of the image). Moreover, the center of smoothed image with high value has lees details and more blurring and this effect on the contrast of the images. Also, the blurred not cover all the image( it is clear when you look to the edge – corner of the image-) because the edge of smoothed image is free from the blurring.
The result of this exercise: the high value of span can effect on the center of the image (blurring) and lost some details of the image. So, the quality of the image reduces.
In this exercise after added the salt and pepper noise to the original image with shot noise strength 3000 the result was a noisy image. After that, the moving average filter and the median filter applied to the noisy image with span (5). So, there is some noise and low contrast image of the filtered image with moving average filter. On the other hand, the filtered image became smoother and better contrast with median filter than the filtered image with the moving average filter.
The result of this exercise: the median filter is better than the moving average filter with span(5). because the median filter is remove the noise and protecting the sharp edges.
In these exercises, the shot strength is increased to 30000 and the result of images have more noise. In this exercise the moving average span has changed to (5, 10 and15) and the median span to (5, 10 and 15) on these images. When the moving average filter has applied with span (5) the filtered image become highlighted and becomes brighter although losses the black objects in some area of the original colors.
Furthermore, when the median filter has applied with span (5) the image still preserving the original value because the function of the median filter used one of the main pixels instead of averaging and the result is new value. Furthermore, by using the moving average filter to the span (10), the resulted image becomes more obscure. In contrary, by using the median filter image with using span (10). The resulted image has less amount of noise and the edge of the black objescts are sharp better than the edges of the images with 5 span. Finally, when applied the moving average filter with the span (15) the
resulted image is blurred while, with median filter and span (15) the resulted image has removed the noise but slightly lost some details of the black objects in the image.
The result of this exercise is that: the resulted image with median filter show more better than the resulted image with moving average filter when the span increased. Because the median filter remove the noise from the image without effect on the original pixels.
Exercise 2.3 :- This exercise show the median filter is ineffective because the amount of noise in the image is high and the median value will be noisy pixel. So that, the median filter is ineffective when the shot strength number increased to high level. On the other hand, the moving average filter show effective because moving average filter can produce the image better than the median filter when the shot strength number increased to high level
The result of this exercise is that: the median filter could be ineffective if the shot strength number increased to high level such as 40500 or more.
Exercise 2.4:- This images show the pit scan with shot noise 30000.and the span values are (25and 50). Moreover, the noisy image was filtered with different filtered ( moving average filter and median filter). The results show that the median filter is also better with this type of images more than moving average filter.
The result of this exercise: the median filter show better than moving average filter when the span value increased while the moving average filter is effective when the shot strength number increased to high level. Moreover, the median filter conserves sharp edges than moving average filter.
Exercise 3.1:- This exercise shows four types of images, original image, noisy image, moving average image and kernel filtered image. The kernel image produced as a result of applied the Gaussian filter on the noisy image. when compare the kernel filtered image to moving average image it is clear to shows that, the kernel filtered image is sharper and has more noise, but not blurring as the filtered images with moving average filter. In addition, when compare the kernel filtered image with other images show that brighter than other images. Furthermore, kernel filtered image has less contrast when compare it with the original image. Moreover, the kernel filtered image is clear and the edges are sharper than the moving average filtered image.
The result of this exercise is that: the user defined kernel filter can improve the image quality and keep image detail and edges than the moving average filter, but the noise is still slightly available with lost some of contrast and some details of image
More Important Things