Abstract :

This Project Presents A CNN-based System For Identifying Safe Landing Zones For Spacecraft On Mars By Analyzing High-resolution Terrain Images. The System Operates In Two Stages. First, It Classifies Each Input Terrain Image As Either "safe" Or "unsafe" Using A Convolutional Neural Network Based On The MobileNetV2 Architecture. Safe Terrain Is Characterized By Flat, Obstaclefree Surfaces, Whereas Unsafe Terrain May Contain Rocks, Craters, Or Steep Slopes. MobileNetV2 Is Chosen For Its Efficiency And High Accuracy In Image Classification, Making It Suitable For Deployment In Resource-constrained Environments Such As Onboard Spacecraft.In The Second Stage, If An Image Is Classified As "safe," The System Further Analyzes It To Identify And Highlight Exact Safe Landing Zones Within The Image. This Is Achieved Through Image Processing Techniques That Detect Flat Regions, Followed By Contour Detection To Locate Continuous Plain Surfaces. These Regions Are Then Validated And Visually Highlighted, Providing Precise Localization Of The Safest Landing Sites. The Combined Approach Ensures Both High-level Terrain Assessment And Finegrained Identification Of Safe Landing Zones, Contributing To Safer Autonomous Planetary Landings. Index Terms — Mars Landing, CNN, MobileNetV2, Image Processing, Safe Landing Zone Detection, Terrain Analysis, Autonomous Navigation.

Published:

29-10-2025

Issue:

Vol. 25 No. 10 (2025)

Page Nos:

380-387

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite