0% Complete
صفحه اصلی
/
سی و دومین کنفرانس بین المللی مهندسی برق
Depth Estimation in Monocular Images of Inside the Sewer Pipes for Navigation of Inspection Robots
نویسندگان :
Zeinab Maroufi
1
Alireza Hadi Hosseinabadi
2
Reza Askari moghadam
3
1- University of Tehran
2- University of Tehran
3- University of Tehran
کلمات کلیدی :
Depth،Estimation،Neural Network،Sewer Pipe
چکیده :
Estimating the depth of a scene from single monocular color images is a fundamental problem in image understanding. Depth estimation is a critical function for robotic tasks such as localization, mapping, and obstacle detection. Recent works based on the development of deep convolutional neural networks provide reliable results. Due to the low cost and relatively small size of monocular cameras, the performance of neural networks for depth estimation from single RGB image has increased significantly. Inspector robots move inside the sewer pipe in an ambiguous environment that has various contaminations and obstacles. As a result of understanding the environment inside the pipe and analyzing the images from the monocular camera, the robot can move more safely and perform the mission more effectively. This paper presents a new deep neural network, called SepiDepth-ASPP. Our approach uses the integration of ASPP and adaptive bins to extract strong global and local contextual features at multiple scales, and then translate them to higher resolutions for clearer depth maps. This network is specially designed for images inside the sewer pipe to more accurately estimate the details of the images in the depth map.This network runs on the dataset inside the sewer pipe and helps the robot comprehend the inside of the pipe environment.
لیست مقالات
لیست مقالات بایگانی شده
A Bidirectional Transformerless Direct AC-AC Dynamic Voltage Restorer with Extended Compensation Range and Up/Down Capability
ُSeyed Mohsen Mortazavi - MohammadHadi Mokhtari - Mohammad Reza Zolghadri
Proposing an indirect distributed approach to apply SSSEP vibrational stimulation
SAHAR SADEGHI - Ali Maleki
Development of Iterative Learning Control Method Based on Markov Parameters for High-Order Discrete-Time Singular Systems
Meysam Azhdari - Tahereh Binazadeh - Ali Gholami
Binomial Distribution based K-means for Graph Partitioning Approach in Partially Reconfigurable Computing system
Zahra Asgari - Maryam Sadat Mastoori
Design and Simulation of Long Slot SIW Leaky Wave Antenna for Automotive Radar Application
Jamal Kazazi - Alireza Rahmani - Mahmoud Kamarei
پیچش زمانی عمیق برای انطباق چندگانه سری های زمانی
سیدعلیرضا نوربخش - نرجس الهدی محمدزاده
A Brief Review on DC-Link Control Strategies in Microgrids
Mehran Seydi - Hassan Moradi CheshmehBeigi - Mohammad Hossein Mousavi
Technical and Economic Analysis of Voltage Sags Mitigation Methods
Sina Shakeri - Javad Khajouei - Saeid Esmaeili
مدیریت بهینه توان در یک ساختمان هوشمند حاوی واحدهای ترکیبی برق و حرارت و منابع تولیدپراکنده در حضور ذخیره ساز انرژی
اسماعیل زحمت کشان
Higher Derivatives Extremum Seeking with Very Slow/ Drifting Sensor
Farzaneh Karimi - Mohsen Mojiri
بیشتر
ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 41.7.4