Peanuts are easily contaminated by various molds during storage and transportation to produce mycotoxins, among which aflatoxin B1(AFB1) is the most common. A novel method for determination of the AFB1 in peanuts based on colorimetric senor technology was proposed. Indicative volatile components of different moldy peanuts were obtained by headspace solid phase microextraction with gas chromatography-mass spectrometry (HS-SPEM-GC-MS). According to the result of the HS-SPEM-GC-MS report, totally 12 kinds of chemical dyes were selected to prepare a color sensitive sensor array with strong specificity, which was used to collect the odor information of peanut samples with different degrees of mildew. Genetic algorithm (GA) combined with back propagation neural network (BPNN) was used to optimize the color component of the preprocessed sensor feature image. Then support vector regression (SVR) was used to construct a quantitative model based on the combination of optimized feature color components to achieve the determination of the AFB1 in peanuts. In this process, the optimization performance of grid search (GS) algorithm and sparrow search algorithm (SSA) on SVR parameter was compared. The results obtained showed that the performance of SSA-SVR model was better than that of GS-SVR model. The correlation coefficients of prediction (RP) of the best SSA-SVR model based on the combination of seven feature color components reached 0.9142. The root mean square error of prediction (RMSEP) was 5.6832μg/kg, and the residual predictive deviation (RPD) value was 2.3926. The overall results demonstrated that it was feasible to use the olfactory visualization technology for the determination of the AFB1 in peanuts. In addition, proper optimization of sensor features and model parameters can further improve the detection performance of the model.