Intellectual Property India Publishes Patent Application for System And Method For Predictive Analysis Of Pedestrian Behavior Using Vision-Based Traffic Scene Interpretation

MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641092340 A) filed by Mrs Kunkulagunta Naga Maha Lakshmi; Mrs Manupala Amani; Mr. E. Prabhakar; Mrs. D. Kavitha; Mr. A. Kiran Kumar Yadav; Ms Vangala Konica Nehal; Mr. Goli Shivaprasad Reddy; Ms. Kota Prathiba; Mr. Dhanavath Manesh; and Mr. Kempu Bharath on July 30, 2026, for System And Method For Predictive Analysis Of Pedestrian Behavior Using Vision-Based Traffic Scene Interpretation.

Inventors include Mrs Kunkulagunta Naga Maha Lakshmi; Mrs Manupala Amani; Mr. E. Prabhakar; Mrs. D. Kavitha; Mr. A. Kiran Kumar Yadav; Ms Vangala Konica Nehal; Mr. Goli Shivaprasad Reddy; Ms. Kota Prathiba; Mr. Dhanavath Manesh; and Mr. Kempu Bharath.

The application for the patent was published on August 07, 2026, under issue no. 32/2026.

Abstract: Abstract The present invention discloses a method and system for real-time pedestrian behavior prediction and collision risk assessment using visual sensing and intelligent learning models. The system utilizes a camera-based input to capture continuous scene data, followed by a Unified Pedestrian Perception Encoder (UPPE) to extract identity-consistent behavioral, spatial, and contextual features. A Trajectory Prediction Network (TPN) with a temporal sequence learning backbone predicts multi-step future pedestrian trajectories, while an Uncertainty-Guided Forecast Refinement (UGFR) module improves prediction stability by selectively refining uncertain trajectory segments. A Predictive Risk Assessment Unit (PRAU) computes collision-related metrics including time-to-collision, trajectory convergence, and probability of interaction. The system further enables adaptive risk evaluation to generate real-time safety alerts based on predicted pedestrian behavior and environmental context. The proposed invention enhances accuracy, robustness, and computational efficiency in dynamic traffic environments, thereby supporting intelligent transportation systems, autonomous driving, and pedestrian safety applications.

Disclaimer: Curated by HT Syndication.