In this paper, a novel, simple, rapid, and low-cost detection device for pesticide residue was constructed. A sensor array based on a cross-responsive mechanism was designed. The data collection and processing system was used to detect fluorescent signal of the sensor arrays, and to extract unique patterns of the tested pesticide residue. Four selected pesticides, carbendazim, diazine, fenvalerate, and pentachloronitrobenzene, were detected by the proposed device. Unsupervised pattern recognition methods, hierarchical cluster analysis and principal component analysis, were used to analyze the data. The results showed that the methods could 100% discriminate the four pesticide residues. According to the standard regression linear curve of the fluorescence intensity and the concentration of pesticide, the quantitative value of the pesticide was detected, and the device obtained responses at concentrations below 8 ppb, and it has a good linear relationship in the range of 0.01-1 ppm. According to the results, the proposed detection device showed excellent selectivity and discrimination ability for the pesticide residues. However, our preliminary study demonstrated that the proposed detection device has excellent potential application for the safety inspection of food.