The Story

The client aimed to implement an innovative solution for the rice industry to measure the quality of rice grains with minimal effort. The primary objective was to streamline the process of obtaining lab results, significantly reducing the time and steps involved in the current, labor-intensive procedure.

Developed a desktop application using Mask R-CNN to scan and analyze rice grains, generating detailed quality reports in 1-2 minutes, improving efficiency and reducing labor-intensive quality control processes.

Amit

Founder & CTO

The challenge

The core challenge was to develop a machine learning model capable of accurately analyzing rice grains from scanned images. The model needed to generate a CSV file with detailed information, including grain height, width, color, location on the image, total count, and other relevant parameters. Additionally, the model had to capture and save individual grain images from the input image.

The solution

To address the challenge, we implemented Mask R-CNN, a deep learning model renowned for its object detection and segmentation capabilities. We fine-tuned the model's parameters, such as thetas, batch size, and epoch cycles, to achieve optimal performance. The solution involved scanning rice grains with a desktop application built on Python, .NET Core 6, .NET Framework 4.8, and SQLite. The generated CSV file, containing detailed metadata, is calculated and displayed within the application, providing comprehensive test reports and maintaining a history of the scans.

The outcome

The implemented solution successfully generates and displays a scanned image report within 1-2 minutes. This rapid processing capability ensures that the rice industry's quality measurement process is significantly more efficient, reducing the time and effort required to obtain accurate lab results.

Planning to outsource software development services?

Contact sales, to start a project, now.