把含有数字的图像,如车牌,身份证等读取到DSP的SDRAM中,利用图像处理算法把数字从图像中定位出来 首先把图像二值化，二值化算法选择固定阈值、直方图或最大类间方差法，比较各个二值化算法的效果；对二值化的图像做边缘提取，选择Sobel或者Laplace边缘提取算法并比较效果；经过二值化和边缘提取后的图像，利用投影法定位数字在图像中的位置，并给出数字在图像中的外接矩形；利用CCS把处理结果显示出来。-To contain digital images, such as license plates and identification cards to read to the DSP' s SDRAM, the use of digital image processing algorithm to locate them from the image　first to image binarization, thresholding algorithm to select a fixed threshold histogram Figure or Otsu method, compare different binarization algorithm results　on binary images to do edge detection, select the Sobel edge detection algorithms or Laplace and compare the results　after binarization and edge images extracted using Location digital projection position in the image and give the number in the image of the bounding rectangle　use of CCS to the results displayed.