Page 66 - tech fest 2025 ב״ש חוברת תקצירים

66
A Smart Assistive Cane for Visual Impairment Using an Embedded Camera-Based System
EEE-A-09
Liron Leizeronok; lizliron@gmail.com Ido Sovorsky; idosoroksky@gmail.com
Advisor: Mr. Amit Twik
SCE - Shamoon College of Engineering, Be’er Sheva
This project presents the development of a smart walking cane for visually impaired individuals, integrating a camera-based obstacle detection system. The system processes images to identify sidewalks, roads, and obstacles, providing users with real-time feedback. The development process included data collection, image preprocessing, model selection, and training with pretrained models. The solution was successfully implemented using only a camera, without relying on additional sensors. The smart cane enhanced mobility, safety, and independence by enabling users to navigate their environment more confidently. The system was tested under various conditions to ensure reliable performance in real-world scenarios.
Keywords: assistive technology, image recognition, mobility enhancement, obstacle detection, smart cane
Data-Log Tool and Parametric Analysis Smart System for ASIC Qualification Tests
EEE-A-10
Adi Abekasis; adiabekasis2@gmail.com
Advisor: Mr. Hanan Ohayon1
1SCE - Shamoon College of Engineering, Be’er-Sheva 1SanDisk Corporation
This project focused on designing an intelligent system for managing and analyzing chip testing in post- silicon validation. The system attained two main objectives: (1) developing a smart platform for chip test management, and (2) demonstrating parameter analysis using machine learning. It can serve as a foundation for a future system capable of predicting chip test results. The system was designed to store and manage tested chip data within a structured database, ensuring accessibility and organization. Additionally, it aggregates data, classifies parameters, and trains a machine-learning model to assist in determining chip functionality. In the future, this capability will be extended to predict the outcomes of untested chips, enhancing efficiency and decision-making in the validation process.
Keywords: automated analysis, data management, machine learning, parameter classification