MUMBAI, India, July 31 -- Intellectual Property India has published a patent application (202411102459 A) filed by Cnh Industrial Technology Services India Private Limited on December 24, 2024, for System And Method Of Determining Orientation Of Cotton Bolls Using Deep Learning.
Inventors include Gupta, Sagar; Lodhi, Vaibhav; Saini, Varun; Kumar, Ashish; Singh, Piyush; Modwel, Garv; and Singh, Maninder.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: SYSTEM AND METHOD OF DETERMINING ORIENTATION OF COTTON BOLLS USING DEEP LEARNING A system (800) and a method of determining orientation of cotton bolls are described. The method comprises obtaining images of cotton bolls growing on cotton plants. The images are pre-processed to obtain refined images having uniform features. A Deep Learning (DL) model (112) predicts location and orientation of the cotton bolls, from the refined images, using anchor boxes of different sizes. The DL model (112) is pretrained on an annotated dataset of cotton bolls which includes a plurality of images of cotton bolls labelled with their orientation. The DL model (112) is pretrained by comparing a predicted output with a ground truth, and the DL model (112) is retrained when a difference between the predicted output and the ground truth exceeds a predefined threshold. (Fig. 1)
Disclaimer: Curated by HT Syndication.
