In-House Crop Image Analysis from Image Capture to AI Inference: Chloros Launches SWALO

Background

In recent years, the volume of field data collected for agricultural research and production management has increased significantly. As a result, there is growing demand for more efficient and standardized crop evaluation processes.

However, implementing AI-based image analysis often requires specialized expertise, training data preparation, and dedicated computing environments, which can create barriers in terms of cost and operational workload. Chloros has previously provided crop image analysis technologies to public agricultural research institutes as well as agrochemical and seed companies. Based on feedback from these users, the company redesigned its solution to make crop image analysis more accessible to a broader range of organizations.

Overview of SWALO

SWALO is an integrated crop image analysis system that combines:

• ・PathfinderClose-range drone flight path generation software
• ・Crop image analysis AI modelsAI models for detecting and analyzing crops and growth conditions from images
• ・SWALO ScannerDetection result aggregation and geospatial mapping

Together, these components provide an integrated environment where users can manage the full workflow—from planning image acquisition to AI inference and result visualization.

This enables more efficient crop evaluation and improved data management across research and agricultural operations.

Enabling In-House Crop Image Analysis

Previously, part of the analysis workflow after image capture was performed by Chloros. With SWALO, the entire process—from image capture to AI inference and result aggregation—can now be completed directly within the user's environment.

The system supports both on-premise and cloud deployment, allowing organizations to operate according to their own data governance policies. In on-premise environments, research data can be analyzed without being transmitted outside the organization, helping protect intellectual property and sensitive research data.

Key Updates

1. Pre-trained AI Models

SWALO introduces pre-trained AI models capable of detecting crops such as rice, wheat, maize, eggplant, strawberry, tomato, and grape.

These models allow users to start using crop image analysis quickly without building models from scratch. The system also supports fine-tuning using small additional datasets, enabling flexible deployment for both research and production management.

Example detections of panicles, fruits, plants, and pest/disease damage using pre-trained models (development image)

Adoption Track Record

Since its commercial launch in 2025, Chloros' crop image analysis technology has been adopted by public agricultural research institutes (2 institutes) and agrochemical and seed companies in Japan and overseas (5 companies). The system has been used in field trials such as crop variety evaluation and agrochemical efficacy testing.

Availability

Strategic Positioning

This update represents Chloros' evolution from providing AI analysis services and tools to delivering an environment where crop image analysis can be executed directly within the customer's infrastructure. By providing a crop image analysis platform usable not only in research and development but also in broader agricultural operations, Chloros aims to support the advancement of data-driven agriculture.