# Deep Learning-Based Image Classification of Pupae from 11 Lepidoptera Pest Species

## TL;DR

This paper shows that deep learning can accurately identify moth pest pupae using images, offering a new tool for pest monitoring in agriculture.

## Contribution

The study introduces a multi-angle image dataset and demonstrates high-accuracy deep learning models for automated pupal pest identification.

## Key findings

- A deep learning model achieved 98.71% accuracy in identifying lepidopteran pest pupae.
- Multi-angle imaging improved model performance and reduced confusion among similar species.
- The dataset and benchmarks provide a foundation for developing field-ready pest monitoring tools.

## Abstract

Traditionally, distinguishing pupae of lepidopteran pests has been challenging due to their subtle morphological differences. To overcome this, we constructed a multi-angle image dataset of pupae from 11 economically important moth pests and tested six state-of-the-art deep learning models for automated identification. The models successfully learned to identify the species with high accuracy—the best reaching over 98% accuracy—confirming that pupal images contain enough visual information for reliable machine-based classification. This work demonstrates a practical path toward developing rapid, image-based tools for pupal pest monitoring in the field, which could significantly improve early detection and management in agriculture.

The morphological identification of lepidopteran pest pupae has long been a difficult task. To explore automated solutions, this study established a standardized, multi-angle image dataset of pupae from 11 economically important lepidopteran pests. We then systematically evaluated six deep learning models, including both convolutional neural networks and Transformer architectures. The results show that all models successfully learned to distinguish the vast majority of species, with Vit-Small achieving the highest accuracy (98.71 ± 0.16%) and the highest F1-score (98.69 ± 0.20%). This confirms that pupal morphology provides sufficient discriminative visual information to support highly accurate automated identification. However, all models exhibited consistent, minor confusion among Helicoverpa armigera, Mythimna separata and Spodoptera exigua. Analysis revealed these errors originated from specific viewing angles of a limited number of specimens, underscoring the value of the multi-angle imaging protocol used in this study. This study transforms pupal identification from a traditional taxonomic difficulty into a solvable computer vision task, providing a dataset, methodological benchmarks, and a feasibility validation for developing image-based tools for pupal-stage pest surveillance.