Venous beading的問題，透過圖書和論文來找解法和答案更準確安心。 我們找到下列包括價格和評價等資訊懶人包

視網膜血管的檢測及其寬度測量的模糊分割方法

為了解決Venous beading的問題，作者NEHAL AHMAD 這樣論述：

Retinal diseases are commonly associated with the abnormalities present in the human eye’s blood vessels that may lead to blindness in childhood. Retinopathy of prematurity (ROP) is one of the most common among these diseases. The regular vascular changes that take place to the posterior pole vesse

ls of a premature baby occur in a predictable manner quickly after birth. These changes are defined as plus disease and are known as abnormal dilation and blood vessel’s tortuosity during ROP that can reach to total retinal detachment. The ophthalmologists now have better understanding of the pathol

ogy and cataract of events that occur in the posterior pole of the eye having ROP. As the vessels generally become thicker and tortuous, the diameter variation is one of the most important concern of vascular proliferative, blindness causing disease such as retinopathy of prematurity plus disease, d

iabetic retinopathy which is venous beading (abnormal variations in diameter along the vein), hypertension and other cardio vascular disease. Various morphological changes take place to retinal blood vessels at different stages with varied conditions of ailments as time passes away. So, preciseness

in the measurement of vessel diameter is a very important part for the effective diagnosis of ROP plus diseases. Therefore, early detection of these morphological changes can help to reduce the probability of blindness occurrence in infants. So, detection of retinal vessels and its width measurement

is the major concern for the rectification of such severe disease and such problems are diagnosed by the proposed method. Firstly, for the detection of the skeleton image of the blood vessels, fuzzy segmentation method was employed as an edge detection technique and for the accuracy comparison of t

he segmentation, Canny’s edge detection algorithm was implemented. And then the skeleton was further utilized for the identification of branching points. For the branching points identification, a contour finding method was used to identify whether each pixel belongs to a bifurcation point, internal

point or an end point. Then, an algorithm was developed to measure the width of the blood vessel in which from the current pixel positions of one edge, the edge pixel of opposite edge was identified by taking perpendicular distance from that pixel. The distance between these two pixels was calculat

ed and the distance between two next pixels and two previous pixels were calculated. The minimum distance among these five pairs of pixels was taken as the width of the vessel. The proposed method was evaluated using the STARE dataset. The performance of the proposed method was compared with existin

g methods where t Snake algorithm was used for the edge detection and Douglas-Peucker algorithm was applied for width measurement. The proposed method outperformed the existing ones. Experimental results indicated that the proposed method was effective to measure the width in comparison to existing

methods.

基於非次取樣輪廓波轉換之視網膜血管分割

為了解決Venous beading的問題，作者辜士哲 這樣論述：

本研究提出一套基於非次取樣輪廓波轉換之視網膜血管分割法，首先對彩色視網膜影像擷取綠色通道，對綠色通道進行非次取樣輪廓波轉換，用線性偵測器在每個尺度下，對非次取樣輪廓波轉換不同方向的subband偵測血管之方向，並擷取線性偵測器偵測到之正交方向，以該正交方向擷取非次取樣輪廓波轉換的係數，並在擷取的係數中計算本研究提出的三種特徵。接著使用支援向量機對視網膜影像中的每個像素進行分類，以上述三種類型的特徵和綠色通道的像素值進行分類，並對分類後得到的二元影像進行後處理，本研究提出三種後處裡的方法，針對視神經盤、可見視野遮罩、非血管區域和雜訊所造成的偵測錯誤進行消除。本研究使用DRIVE資料庫進行效能評

估，其準確率為0.9413。