論文簡介

      

      

       

 


 

       

 

 

近年來,多數的研究已經說明在仿細胞神經網路(Cellular Neural Networks; CNN) 型態的架構下,提供一個可用程式化的方式來處理多數複雜的影像處理工作。CNN的架構中包含了可做即時處理的平行類比計算單元,其中有一個理想的特性是這些處理單元是有規則的二維陣列排列,且本身與鄰近的細胞單元為區域性的元件連接。由於此種特性,使得這種架構很容易在超大型積體電路上實現。因此在本論文中提出以CNN為基礎的紋理邊界偵測之新的影像處理系統與它的類比電路實現

本論文所提出的紋理邊界偵測技術,是模仿人類眼球表面層上的結構行為來偵測影像的紋理邊界。利用多數且平行CNN處理器計算技術的創新,取代以往複雜的數位式影像紋理邊界偵測。對於即時運算方面,它被設計成以CNN為基礎的架構,可以用平行即時處理的類比式電路來實現,大大地增加其執行的效率。而CNN的設計電路採多層次 (Multi-layer) 的方式,以5×5為基礎的細胞核心,將處理影像大小擴展成32×32處理陣列。同時為了降低電路複雜度,採用電流模式 (Current mirror;電流鏡) 的設計架構,且延伸成為可正負雙向電流導通,更容易來實現每個神經細胞的權重比例 (即電流增益),也使得在節點上的多數訊號易於結合。由於CNN具有陣列式平行處理和區域性的元件連接特性,因此很適合實現於混合訊號標準的CMOS製程上。

 


 

       

 

Block diagram

It consists of the 16´16 CNN array with templates, the analog absolute value circuit, and the summation unit. To reduce design complexity and die size for sophisticated process technology in our experiment, a programmable current-mode CNN array is designed, and every current-mode CNN array can be replace by the programmable current-mode CNN array with different template A, B and threshold I.

At first, a digital image is fed into the network as input values in CNN array. These currents are defined positive and assigned to the positive part of the CNN sigmoid function. The network then performs the Gabor filters with four orientations (four kinds of template A). Here we will obtain the results with both positive and negative values in the steady state and the results are fed into the analog absolute value circuits. After that, the results become all positive values and fed into Gaussian filters with the same templates. The results of the Gaussian filters are all positive values and then fed into distance units with the same templates. We will feed results of distance units into analog absolute value circuits one more, and get positive results. The network then performs the summation units implemented by connecting results of five absolute value units pixel by pixel to new results of a new array (because of current-mode). Finally, the network feeds the results of summation units and a threshold value Ith which is the mean of these results to threshold units with the same templates.

 

 


 

       

 

A CNN cell

 


 

 

(a)

(b)

(d)

(c)

(e)

Fig. 1: This demonstrates the results of both modified and CNN-based TBD.(a)input; (b) and (c) show the results before and after threshold processing of modified TBD. (d) and (e) show the results before and after threshold processing of CNN-based TBD.

 


 

(a)

(b)

(d)

(c)

(e)

Fig. 2: This demonstrates the results of both modified and CNN-based TBD.(a)input 2; (b) and (c) show the results before and after threshold processing of modified TBD. (d) and (e) show the results before and after threshold processing of CNN-based TBD.

 


 

(a)

(b)

(d)

(c)

(e)

Fig. 3: This demonstrates the results of both modified and CNN-based TBD.(a)input 3; (b) and (c) show the results before and after threshold processing of modified TBD. (d) and (e) show the results before and after threshold processing of CNN-based TBD.

 


 

(a)

(d)          (e)

(f)

(b)

(c)

Fig. 4: This demonstrates the results of CNN-based TBD.(a)input; (b) and (c) show the results before and after threshold processing. (d) shows the results of Gabor filters, and (e) shows results of rectifier processing after (d). (e) shows the results of Gaussian filters

 


 

(a)

(b)

(c)

Fig. 5: This demonstrates the 16×16 results of Gaussian filter for CNN circuits with using the input as shown in Fig. 4 (a). The results of five channels after Gaussian filter is shown in (a), the result after distance processing is shown in (b), and the result is shown in (c).