Abstract:Qualitative analysis with near-infrared and mid-infrared spectroscopy and extreme learning machine algorithm were used for discriminate sesame oil from adulterated sesame oil with soybean oil. Analyzing 90 adul－ terated samples which were prepared with sesame oil， soybean oil of different brands and batches and 40 sesame oil samples detected the adulteration.Set network parameters a limit of extreme learning machine algorithms， and compared the result of near-infrared and mid-infrared spectroscopy， and studied the feasibility of the spec－ trum detection method and classification accuracy， in order to make basis of theory and practical for the sesame oil adulteration of spectrum detection technology fast discrimination analysis， inhibition of sesame oil adulter－ ation happening.