Desige Of The PCNN Image Segmentation System Based On Nios Ⅱ

This thesis mainly focuses on the realization of PCNN image segmentation algorithm on NiosⅡ. and it takes the advanced criterion of combining minimum cross-entropy with OTSU as the image segmentation part. This system can complete the image segmentation quickly and accurately. Compared with traditional CMOS, it has advantage of less occupancy, short development cycle. The whole system of image segmentation based on NiosⅡconsists of four parts, which are introduced simply as below:1. PCNN image segmentation moduleThis is the core of design which is about the realization of PCNN image segmentation algorithm on NiosⅡ. Combining minimum cross-entropy of binary image with OTSU as the criterion to measure the best cycle index.The image segmentation can be carried out automatically; this can achieve a much better effect.2. Data receiving moduleIt takes the Ethernet to communicate and transmit the collected images. Centering on DM9000A chip, it needs to use theμCOS-Ⅱoperation system integrated on NiosⅡprocessor to control. Thus, the images being waited for segmentation can be received from different host machines.3. Storage module of segmented imagesThe original images and the image data witch after segmentation can be storied on SD card. when the segmenting begin, the data on SD card can be directly input into PCNN image segmentation module for processing. After segmentation, the image data can also be storied.This fully uses SD card's advantages of cheapness and compatibility with different platforms.4. Image displaying moduleThis module is divided into two parts:VGA displaying and TFT-LCD. The original images and segmented binary images will be displayed on TFT-LCD and VGA-LCD, which is convenient to compare the segmentation effects. The program of TFT-LCD driver, for expending the function of Altera DE2-70 board, is mainly consisted of scanty initialization, status register initialization, read and write control.