MUMBAI, India, June 24 -- Intellectual Property India has published a patent application (202411093817 A) filed by Agnext Technologies Private Limited on November 29, 2024, for System And Method For Precise Grain Segmentation And Analysis Using Dual Scanners.
Inventors include Subrat Kumar Panda; Yoginder Kumar; Shashi Praksh; Hemant Kumar Choudhary; Ashish Pati; and Mandeep Saini.
The application for the patent was published on June 12, 2026, under issue no. 24/2026.
Abstract: The present disclosure relates to a system 100 and method for precise grain segmentation and analysis using dual scanners. The system employs a lower scanner 102 and an upper scanner 104 to capture synchronized dual-view images of grain samples 106. A convolutional neural network–based background elimination module 110 removes non-grain pixels and generates segmented grain masks, followed by a cluster detection filter 112 and an edge detection model 114 to isolate individual grain contours. A calibration module 116 computes a homography transformation used by a grain mapping module 118 to establish accurate correspondence between lower- view and upper-view grain images. A physical parameters computation module 120 determines grain length, width, area, and shape ratios, which are further analyzed by an AI classification model 122 to generate class labels and confidence scores. Results are processed through a results generation module 130 and stored in a storage database 132, enabling automated and explainable grain-quality assessment. FIG. 1 will be the reference figure.
Disclaimer: Curated by HT Syndication.
