Vegetation coverage is a widely used parameter to measure global and regional environment change. Evaluating the accuracy and efficiency of vegetation coverage using digital photography under various survey points' densities and distribution patterns has an important referential significance for providing an optimized field measurement method. The vegetation field measurement was carried out in a sample with Artemisia ordosica shrubs in Mu Us sandy land using vertical hoisting digital camera, with four densities and nine distribution patterns of survey points. The results showed that: different density of survey points led to a slight accuracy difference, and the precision improves as the density increases. The sample size had great impact on the precision. Different point distribution patterns led to significantly different results. "Diamond" pattern can get relatively higher degree of accuracy with least points and shortest walking distance in field survey. It's the best choice that could meet the requirements of the maximum precision and minimum workload in the vegetation field measurement.