The invention relates to a desert plant remote sensing identification method using object-oriented multi-source information fusion. The method comprises the following five units: a large-area multi-source information inlay unit, an object-oriented image segmentation unit, a multi-source information fusion unit, an interpretation rule making unit and a desert plant identification unit. In the prior art, the following conditions that desert plant distribution is sparse; growth spectrum information is weak; desert plants are easily influenced by diffuse reflection; and phenomena of a mixed pixel and a same object with different spectrums are remarkable are not considered. Aiming at the above problems, the desert plant remote sensing identification method using the object-oriented multi-source information fusion is researched and developed so that problems that desert plant types and space-time distribution can not be accurately identified in fields of ecological remediation, a climatic change, a biological diversity, genetic development, poverty elimination and the like can be solved. A technical support is provided for a correlation research mechanism, a commonweal organization, an enterprise and a government. A data analysis basis is provided for exploitation, construction and management of a desert plant industry in an arid region.