Optoelectronic wingbeat signatures of Tephritid fruit flies and non-target species: A dataset for automated pest monitoring

This repository contains the dataset and hardware design for the study: "Automated bioacoustic monitoring of Anastrepha fraterculus and Ceratitis capitata: Optoelectronic sensor design, data acquisition protocols, and Machine Learning classification". The dataset includes 2,688 segmented wingbeat events acquired using an original optoelectronic sensor integrated into commercial McPhail traps. Data collection involved two target pests (Anastrepha fraterculus and Ceratitis capitata) and two non-target species (Hermetia illucens and Drosophila suzukii) to evaluate classifier selectivity. Data Structure: The samples are partitioned into Training (70%), Validation (15%), and Testing (15%) sets as follows: * Anastrepha fraterculus: 924 samples. * Ceratitis capitata: 924 samples. * Hermetia illucens: 738 samples. * Drosophila suzukii: 102 samples. Files included: 1. Dataset_Bioacoustic_FruitFlies.zip: Segmented audio events in .wav format (22.05 kHz, 32-bit float). 2. Sensor_Hardware_Design.zip: Technical drawings (PDF) and electronic schematics (PDF) for the optoelectronic sensor integrated with McPhail traps. 3. README.txt: Documentation regarding dataset acquisition, hardware specifications, and signal processing protocols.