Successful forest resources monitoring requires an effective method to collect information related to forest area and tree crown.Current field-based assessment methodology provides the needed information

but is costly

and therefore assessment frequency is low.High-resolution remote sensing techniques provide a potentially low-cost alternative to field-based assessment.This paper proposed a method of monitoring survival rate and growth condition of economic forests in the farmland being returned to forests by using high resolution Quickbird imagery in Zhangjiakou

Hebei province.Multi-resolution image segmentation and object-oriented image analysis approaches were used for extracting farmland being returned to forests and tree crown information from Quickbird data.Firstly

tree crown classifictation was conducted using object-oriented image analysis

then

a program was developed to acquire tree crown factors and survival rate the classified from tree crown map.Finally

accuracy assessment was done according to field survey data

the results showed that the average error of tree crown size was as follows: 0.337m in east-west

0.433m in south-north

the accuracy of survival rate was 89.837%.The study provides a scientific basis for planning

management and decision support of the project of returning farmland to forests.