Leaf wilting is a common phenomenon in plants. Wilting detection is of high importance for plant precision management. In this study, the 3D image of plant leaf was collected using a timeofflight (TOF) camera to quantificationally identify the degree of plant leaf wilting. The 3D information of targeted leaf was examined by the custom wilting index based on discrete time Fourier transform, in which the leaf is equivalent to a mathematical surface in 3D space. Rapid dehydration and rehydration method was used to make plant leaf state different wilting degrees. The image acquisition experiment was proceeded by a acquisition system based on the seeding of cucumber and zucchini. De-noising, threshold segmentation and dot product were used in image preprocessing after the images were collected, and then the wilting index was calculated. The experiments were divided into three periods, which means, dehydration period Ⅰ, rehydration period and dehydration period Ⅱ, and the trend of wilting index was favorable at each stage. The universality was simulated through the variance analysis of a piece of leaf and different pieces of leaf from the same plant. The significance levels are respectively 0.8904, 0.2922 and 0.9036, which are all larger than 0.05. In general, the feasibility and accuracy of the identification of leaf wilting using TOF camera are verified. As leaf wilting embodys in many aspects because of different kinds of factors, control variate method was used in this paper. The identification of outdoor plant leat wilting needs to be intensively studied.