Do you want to publish a course? Click here

Roads Extraction from Satellite Images using Convolution Neural Network Model (Deeplabv3+) A Case Study in Lattakia city

استخراج الطرق من صور الأقمار الصناعية باستخدام نموذج الشبكة العصبونية الالتفافية (+Deeplabv3): حالة دراسية في مدينة اللاذقية

1172   3   0   0.0 ( 0 )
 Publication date 2023
and research's language is العربية
 Created by Shamra Editor




Ask ChatGPT about the research

The purpose of this paper is to extract roads from satellite images, based on developing the performance of the deep convolutional neural network model (Deeplabv3+) for roads segmentation, and to evaluate and test the performance of this model after training on our data.This experimental study was applied at Google Colab cloud platform, by software instructions and advanced libraries in the Python.We conducted data pre -processing to prepare ground truth masks,then we trained the model.The training and validation process required (Epochs=4), by(Patch Size=4images).The Loss function decreased to its minimum value (0.025). Training time was three hours and ten minutes, aided by the advanced Graphics Processing Unit (GPU) and additional RAM.We achieved good results in evaluating the accuracy of the predictions of the trained model (IoU = 0.953). It was tested on two different areas, one of which is residential and the other agricultural in Lattakia city. The results showed that the trained model (DeepLabv3+) in our research can extract the road network accurately and effectively.But its performance is poor in some areas which includes tree shadows on the edges of the road, and where the spectral characteristics are similar to the road, such as the roofs of some buildings, and it is invalid for extracting side and unpaved roads. The research presented several recommendations to improve the performance of the (Deeplabv3+) in extracting roads from high-resolution satellite images, which is useful for updating road maps and urban planning works.


Artificial intelligence review:
Research summary
يهدف هذا البحث إلى استخراج الطرق من صور الأقمار الصناعية باستخدام نموذج الشبكة العصبونية الالتفافية العميقة (Deeplabv3+). تم تطبيق الدراسة على منصة غوغل كولاب باستخدام مكتبات بايثون المتطورة. تضمنت العملية معالجة مسبقة للبيانات، تدريب النموذج، والتحقق من أدائه. استغرقت عملية التدريب ثلاث ساعات وعشرة دقائق باستخدام وحدة معالجة الرسومات المتطورة (GPU) وذاكرة وصول عشوائي إضافية. أظهرت النتائج دقة عالية في استخراج الطرق (IoU = 0.953)، ولكن النموذج واجه صعوبات في بعض المناطق مثل تلك التي تحتوي على ظلال الأشجار أو خصائص طيفية مشابهة للطرق. قدم البحث توصيات لتحسين أداء النموذج في استخراج الطرق من صور الأقمار الصناعية عالية الدقة، مما يسهم في تحديث خرائط الطرق وأعمال التخطيط الحضري.
Critical review
دراسة نقدية: على الرغم من أن البحث قدم نموذجًا فعالًا لاستخراج الطرق من صور الأقمار الصناعية، إلا أن هناك بعض النقاط التي يمكن تحسينها. أولاً، النموذج يواجه صعوبات في التعامل مع الظلال والأجسام ذات الخصائص الطيفية المشابهة للطرق، مما يشير إلى ضرورة تحسين معالجة البيانات المسبقة أو تعديل النموذج ليكون أكثر دقة في هذه الحالات. ثانياً، لم يتم اختبار النموذج على نطاق واسع من البيانات المتنوعة، مما قد يؤثر على تعميم النتائج. أخيرًا، يمكن أن تكون هناك حاجة إلى تحسينات إضافية لاستخراج الطرق الفرعية وغير المعبدة بشكل أكثر دقة.
Questions related to the research
  1. ما هو الهدف الرئيسي من البحث؟

    الهدف الرئيسي هو استخراج الطرق من صور الأقمار الصناعية باستخدام نموذج الشبكة العصبونية الالتفافية العميقة (Deeplabv3+).

  2. ما هي المنصة المستخدمة لتطبيق الدراسة؟

    تم تطبيق الدراسة على منصة غوغل كولاب باستخدام مكتبات بايثون المتطورة.

  3. ما هي دقة النموذج في استخراج الطرق؟

    حقق النموذج دقة عالية في استخراج الطرق بقيمة (IoU = 0.953).

  4. ما هي التحديات التي واجهها النموذج في استخراج الطرق؟

    واجه النموذج صعوبات في استخراج الطرق في المناطق التي تحتوي على ظلال الأشجار أو خصائص طيفية مشابهة للطرق، وكذلك في استخراج الطرق الفرعية وغير المعبدة.


References used
The purpose of this paper is to extract roads from satellite images, based on developing the performance of the deep convolutional neural network model (Deeplabv3+) forroads segmentation, and to evaluate and test the performance of this model after training on our data.This experimental study was applied atGoogle Colab cloud platform, by software instructions and advanced libraries in the Python.We conducted data pre -processing to prepare ground truth masks,thenwe trained the model.Thetraining and validation process required (Epochs=4), by(Patch Size=4images).The Loss function decreased to its minimum value (0.025). Training time was three hours and ten minutes, aided by the advanced Graphics Processing Unit (GPU) and additional RAM.We achieved good results in evaluating the accuracy of the predictions of the trained model (IoU = 0.953). It was tested on two different areas, one of which is residential and the other agricultural in Lattakia city. The results showed that the trained model (DeepLabv3+) in our research can extract the road network accurately and effectively.But its performance is poor in some areas which includes tree shadows on the edges of the road, and where the spectral characteristics are similar to the road, such as the roofs of some buildings, and it is invalid for extracting side and unpaved roads. The research presented several recommendations to improve the performance of the (Deeplabv3+) in extracting roads from high-resolution satellite images, which is useful for updating road maps and urban planning works.
Christopher, S.;Christopher, H. DeepLearningNeuralNetworks forLandUse LandCoverMapping. IGARSS -IEEE International Geoscience and Remote Sensing Symposium, 2018, pp. 2995–2990
A Beginner’s Guide to Segmentation in Satellite Images: Walking through Machine Learning Techniques for Image Segmentation and Applying Them to Satellite Imagery. https://www.gsitechnology.com/Beginners-Guide-to-Segmentation-in-Satellite-Images(Accessed95-92-2022)
Chen, L.; Qianli, Z.; Papandreou, G.; Schroff, F.; Adam,H.Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation, Computer Vision –ECCV, 2018, pp 833–851
Darwishe, D.; Mohammad, A.; Chaaban, F. Developing a Model of Deep Learning by ANNs for Urban Areas Extraction from Remote Sensing Images -Study Area: HomsTartous, Al-Baath University Journal, V. 43, NO. 7, 2021, PP.11-42
rate research

Read More

In this paper, we present a new algorithm to automate the detection and extraction of buildings from satellite images, this algorithm is distinguished since it overcomes some obstacles that limit detecting within other methods, such as the differe nce in shape, color, and height of buildings, and it doesn't need multi-spectral images or other complex and high cost images.
This study constitutes a preliminary step to develop a mathematical model for predicting traffic accidents in the city of Lattakia, based on a number of external factors, which include engineering characteristics, traffic incursions, and traffic acci dent data. As for its main goal, it is to reduce the number of traffic accidents expected in the future on the main streets in the city, as the study was conducted on various arterial streets in them in terms of their importance and in terms of the number of traffic accidents recorded on them, and in terms of the diversity of their engineering characteristics, in order to have sufficient familiarity with the traffic conditions in The city for various reasons, does not depend on the human behavior of the drivers or on the characteristics of the vehicle. A statistical analysis of traffic accident data for the years 2014, 2015, 2016 and 2017 was conducted on urban streets in Lattakia, where accidents were classified according to their severity, time of occurrence and place of their occurrence, and the necessary data were collected and digitized within a software environment in Microsoft Excel, and then a model was built Predicting the use of the artificial neural networks tool in the MATLAB program, in which data for 319 traffic accidents that were recorded in the years 2015, 2016 and 2017, were entered, which were divided into three groups (training, validation and testing). The structural neural network (10-10-1) gave high values ​​of the correlation coefficient, as the total R value during the three stages was 0.931236, which is very close to one, and therefore the designed network is ideal and achieves the response to predict traffic accidents monthly with very high accuracy.
يهدف البحث إلى تقديم دراسة مرجعيّة مفصلة عن استخدام الشبكات العصبونية الإلتفافية (CNNs) في استخراج الميزات (Features) من الصور. وسيتطرق البحث إلى التعريف بمعنى الميزات (Features) الخاصة بالصور وأهميتها في تطبيقات معالجة الصورة. وسيتم أيضاً التعريف بالشبكات العصبونية الإلتفافية (CNNs) وبنيتها و طريقة عملها وأنواع المقاربات والمنهجيات المستخدمة في تدريبها لاستخراج الميزات (Features) من الصور.
A new face detection system is presented. The system combines several techniques for face detection to achieve better detection rates, a skin colormodel based on RGB color space is built and used to detect skin regions. The detected skin regions are the face candidate regions. Neural network is used and trained with training set of faces and non-faces that projected into subspace by principal component analysis technique. we have added two modifications for the classical use of neural networks in face detection. First, the neural network tests only the face candidate regions for faces, so the search space is reduced. Second, the window size used by the neural network in scanning the input image is adaptive and depends on the size of the face candidate region. This enables the face detection system to detect faces with any size.
In recent years, the problem of classifying objects in images has increased by using deep learning as a result of the industrial sector requirements. Despite of many algorithms used in this field, such as Deep Learning Neural Network DNN and Convolut ional Neural Network CNN, the proposed systems to address this problem Lack of comprehensive solution to the difficulties of long training time and floating memory during the training process, low rating classification. Convolutional Neural Networks (CNNs), which are the most used algorithms for this task, were a mathematical pattern for analyzing images data. A new deep-traversal network pattern was proposed to solve the above problems. The aim of the research is to demonstrate the performance of the recognition system using CNNs networks on the available memory and training time by adapting appropriate variables for the bypass network. The database used in this research is CIFAR10, which consists of 60000 colorful images belonging to ten categories, as every 6,000 images are for a class of these items. Where there are 50,000 training images and 10,000 test tubes. When tested on a sample of selected images from the CIFAR10 database, the model achieved a rating classification of 98.87%.

suggested questions

comments
Fetching comments Fetching comments
mircosoft-partner

هل ترغب بارسال اشعارات عن اخر التحديثات في شمرا-اكاديميا