It is difficult to detect plant leaf edge in green house environment because of noise and incompletion. For this problem, an edge detection based on fuzzy logic was proposed. Firstly, according to the statistics characteristic of neighbor pixels and color difference between plant and soil, membership function of noise and edge was defined by using linear function. Secondly, fuzzy rules were deduced by using Sugeno fuzzy model, so as to decrease the rule amounts. Then, with the results of fuzzy reasoning, fuzzy filter and fuzzy edge detection were performed. The experiment results indicated that the proposed method could overcome the problem of large amounts of false edges which caused by color difference between leaves and backgrounds in conventional algorithms. The proposed algorithm could detect edges in noisy environment effectively and could be used for the plant leaf edge detection in greenhouse monitoring. 