An integrated sensor array system is developed for detection and identification of environmental pollutants in diesel and gasoline exhaust fumes. The system includes a low noise floor analog front-end followed by a signal processing stage. Classification methods are used since the pollutants are often encountered as complex mixtures. In this paper, we present techniques to detect, digitize and classify analytes. This is done by extracting appropriate features from sensor data and using pattern recognition methods to identify the analytes. The detection analog front-end achieves 54dB SNR. The low-noise digitization technique is presented along with the feature extraction and classification algorithms. Comparative results are given for a series of pattern classifiers.