The client’s existing desktop-based application for plant phenotyping was incomplete. It was told that the existing IT Vendor could not improve the app from where it stood then. Manual image segmentation was time-consuming and often led to inaccuracies. They also faced challenges in managing and maintaining their data archives.

Experion embarked on a transformative journey to address these challenges. Leveraging our expertise in data science, image processing, and Computer Vision, we proposed a comprehensive solution.

Re-Architecture and Deep Learning: Our team of Data Scientists re-architected and rebuilt the existing application. We incorporated a cutting-edge deep learning-based segmentation model, which was deployed on the AWS cloud infrastructure. This model enabled highly precise background removal and the measurement of image segments.

Efficient Data Processing: The new system allowed the client to analyze five unique crop types in under a minute for most images, a significant improvement over their previous setup. This achievement was facilitated by the deep learning model’s ability to process images swiftly and accurately.

Statistical Output: The system generated statistical files from the analyzed data, which were then used for further research, reporting, and analysis. This output provided valuable insights into crop health, growth patterns, and potential issues, aiding in research and decision-making.

Scalable Cloud Environment: We implemented a scalable cloud-based environment, allowing the client to run multiple instances of image analysis simultaneously. This increased the throughput and efficiency of their operations.

Automation: To streamline their workflow, we automated tasks such as archiving old images and routine maintenance of the application and storage files. This reduced the administrative overhead and ensured the system’s reliability.

The implementation of our data-driven plant phenotyping application transformation yielded significant outcomes:

• Remarkable Efficiency: The client achieved a 100-fold reduction in image processing time, allowing them to analyze large datasets swiftly and accurately.
• Scalability: The cloud-based environment enabled the client to scale their operations as needed, accommodating increased workloads and data volumes efficiently.
• Enhanced Accuracy: The deep learning-based segmentation model significantly improved the accuracy of image analysis, providing more reliable research insights.

Experion’s collaboration with the client resulted in a state-of-the-art plant phenotyping application, revolutionizing their research capabilities. By harnessing the power of deep learning and cloud computing, the client now possesses a versatile, efficient, and scalable solution that accelerates their research and enhances their contribution to global agriculture.