This survey provides a comprehensive review of image segmentation techniques using deep learning, covering a wide range of methods and applications. It begins by introducing the importance of image segmentation in various fields such as scene understanding, medical image analysis, and augmented reality. The survey then delves into the development of deep learning models for image segmentation, highlighting the success of these models in achieving high accuracy on popular benchmarks. The survey is organized into several sections, each focusing on different aspects of deep learning-based image segmentation: 1. **Overview of Deep Neural Networks**: This section covers the fundamental deep learning architectures, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), long short-term memory (LSTM) networks, encoder-decoder models, and generative adversarial networks (GANs). 2. **DL-Based Image Segmentation Models**: This section reviews over 100 deep learning-based segmentation methods, categorized into 10 groups based on their architectural contributions. These methods include fully convolutional networks, convolutional models with graphical models, encoder-decoder models, multi-scale and pyramid network-based models, R-CNN-based models, dilated convolutional models, recurrent neural network-based models, attention-based models, generative models, and CNN models with active contour models. 3. **Popular Datasets**: The survey discusses around 20 popular image segmentation datasets, categorized into 2D, 2.5D (RGB-D), and 3D images, providing insights into their characteristics and usage. 4. **Evaluation Metrics and Performance**: This section presents popular metrics for evaluating deep-learning-based segmentation models and reports the quantitative results and experimental performance of the reviewed methods. 5. **Challenges and Future Directions**: The survey concludes with a discussion of the main challenges and potential future research directions in deep learning-based image segmentation. The survey aims to provide a comprehensive understanding of the current state of deep learning-based image segmentation, highlighting the strengths, limitations, and future opportunities in this field.This survey provides a comprehensive review of image segmentation techniques using deep learning, covering a wide range of methods and applications. It begins by introducing the importance of image segmentation in various fields such as scene understanding, medical image analysis, and augmented reality. The survey then delves into the development of deep learning models for image segmentation, highlighting the success of these models in achieving high accuracy on popular benchmarks. The survey is organized into several sections, each focusing on different aspects of deep learning-based image segmentation: 1. **Overview of Deep Neural Networks**: This section covers the fundamental deep learning architectures, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), long short-term memory (LSTM) networks, encoder-decoder models, and generative adversarial networks (GANs). 2. **DL-Based Image Segmentation Models**: This section reviews over 100 deep learning-based segmentation methods, categorized into 10 groups based on their architectural contributions. These methods include fully convolutional networks, convolutional models with graphical models, encoder-decoder models, multi-scale and pyramid network-based models, R-CNN-based models, dilated convolutional models, recurrent neural network-based models, attention-based models, generative models, and CNN models with active contour models. 3. **Popular Datasets**: The survey discusses around 20 popular image segmentation datasets, categorized into 2D, 2.5D (RGB-D), and 3D images, providing insights into their characteristics and usage. 4. **Evaluation Metrics and Performance**: This section presents popular metrics for evaluating deep-learning-based segmentation models and reports the quantitative results and experimental performance of the reviewed methods. 5. **Challenges and Future Directions**: The survey concludes with a discussion of the main challenges and potential future research directions in deep learning-based image segmentation. The survey aims to provide a comprehensive understanding of the current state of deep learning-based image segmentation, highlighting the strengths, limitations, and future opportunities in this field.