Is this an image stack of NDVI images from which you need to extract the minimum and maximum NDVI for each pixel? If so, you could try the Min/Max Value image tool available in the Code Library. If these are single band NDVI images, you can just calculate statistics to find the min/max of the band.

It is NDVI image. I want to acquire the fraction of vegetation cover from the NDVI image. It is ok to find the min\max value by using the calculate statistics. But I want to look for the value of bare soil and full dense vegetation to replace the min\max value. NDVIsoil and NDVIfull dense vegetation are assumed to be constants and were extracted from the histogram of the NDVI in the whole study area. But there are also water body and snow cover area which NDVI value are minus. Should I mask the water body and snow cover area at first, and then use the band statistics? Somebody said a non-Gaussian but a broad shape was found for dense vegetation cover and NDVIfull vegetation cover corresponds to the 1% maximum percentile in the histogram of NDVI. Maybe it is can work. But I have no idea how to get the value of 1% maximum percentilein the histogram of NDVI. Do you know how to get it? Thank you very much:)

Yes, you could use a mask, or just an ROI of your areas of interest (vegetation, soil) and use this when calculating statistics as a 'stats subset'.

Thank you very much:) You are very warm hearted.