05_Segmentation_P2

Image Processing Image Processing Segmentation 1 Process of partitioning a digital image into multiple segments (sets of pixels) 2 Clustering pixels into salient image regions, i e , regions correspon[.]

Image Processing Segmentation Process of partitioning a digital image into multiple segments (sets of pixels) Clustering pixels into salient image regions, i.e., regions corresponding to individual surfaces, objects, or natural parts of objects Image Processing Segmentation Used to locate objects and boundaries (lines, curves, etc.) in images Process of assigning a label to every pixel in an image such that pixels with the same label share certain visual characteristics Image Processing Segmentation Two of the most common techniques: thresholding and edge finding Image Processing Segmentation Two of the most common techniques: edge finding thresholding Image Processing A Threshold The original image: The objects in image: A parameter θ called the brightness threshold is chosen and applied to the image f(x, y) as follows: (x, y) ∈ Object or ⇔ f(x, y) ≥ θ (x, y) ∈ Object ⇔ f(x, y) ≤ θ Image Processing A Threshold The original image: The objects in image: Remark: The output is the label "object" or "background" which, due to its dichotomous nature, can be represented as a Boolean variable "1" or "0" Image Processing A Threshold The original image: The objects in image: How to choose the threshold θ ? Image Processing Fixed threshold A threshold will be chosen independently of the image data Image Processing Fixed threshold Source and segmented images with fixed threshold 128 Image Processing Ảnh I: Ảnh I phân đoạn với ngưỡng 200 cho thành phần RED: θ R = 200 Isodata algorithm m1 = 0; m2 = L; teta = (m1 + m2) / stop = false while !stop ts1 = 0; ts2 = ms1 = 0; ms2 = for i = to teta ts1 = ts1 + h(i) * i ms1 = ms1 + h(i) m1 = ts1/ms1 for i = teta to L ts2 = ts2 + h(i) * i ms2 = ms2 + h(i) m2 = ts2/ms2 tg = Round((m1 + m2) / 2) if teta - tg  < ε stop = true teta = tg loop θ = 116, k=4 Image Processing Triangle algorithm A line is constructed between the maximum of the histogram at brightness bmax and the lowest value bmin = (p=0)% in the image Image Processing Triangle algorithm The distance d between the line and the histogram h[b] is computed for all values of b from b = bmin to b = bmax Image Processing Triangle algorithm The brightness value bo where the distance between h[bo] and the line is maximal is the threshold value, that is, θ= bo Image Processing Triangle algorithm Source and segmented images with threshold 152 chosen by triangle algorithm Image Processing Background-symmetry algorithm dominant peak (183) Assumes a distinct and dominant peak for the background that is symmetric about its maximum Image Processing Background-symmetry algorithm •The maximum peak (maxp) is found by searching for the maximum value in the histogram •Searching on the non-object pixel side of that maximum to find a p% point Image Processing Background-symmetry algorithm the object pixels are located to the left of the background peak at brightness 183, that mean h(183) = max {h(a): ≤ h(a) ≤ 255 } = 351 search on the right of that peak to locate to find 95% The total number of pixels in the image is 17424 and the total number of pixels on the right of peak is 8241, about 95% (94.59%) of 17424/2 = 8712 Image Processing Background-symmetry algorithm At which brightness value 5% of the pixels lie to the right (are above)? This occurs at brightness 216 The number of pixels on the right of 216 is 936, equal to 5% (0.0537) the total number of pixels in the image: 17424 Because of the assumed symmetry, we use as a threshold a displacement to the left of the maximum that is equal to the displacement to the right where the p% is found