This survey provides a comprehensive review of deep learning-based image segmentation methods, covering a wide range of approaches including fully convolutional networks, encoder-decoder models, multi-scale and pyramid networks, recurrent neural networks, attention-based models, and generative models. The paper discusses the similarities, strengths, and challenges of these models, examines widely used datasets, and reports performance metrics. It also explores promising future research directions in the field. The survey categorizes deep learning-based works into 10 main categories based on their technical contributions. These include fully convolutional networks, convolutional models with graphical models, encoder-decoder models, multi-scale and pyramid network-based models, R-CNN-based models for instance segmentation, dilated convolutional models and DeepLab family, recurrent neural network-based models, attention-based models, generative models and adversarial training, and CNN models with active contour models. The paper provides an overview of around 20 popular image segmentation datasets, grouped into 2D, 2.5D (RGB-D), and 3D images. It also presents a comparative summary of the properties and performance of the reviewed methods on popular benchmarks. The paper discusses the main challenges and future directions for deep learning-based image segmentation, including the need for more efficient models, better handling of context, and improved performance on complex tasks. It also highlights the importance of transfer learning and the use of pre-trained models for segmentation tasks. The survey concludes with a discussion of the current state of the field and the potential for future advancements in deep learning-based image segmentation.This survey provides a comprehensive review of deep learning-based image segmentation methods, covering a wide range of approaches including fully convolutional networks, encoder-decoder models, multi-scale and pyramid networks, recurrent neural networks, attention-based models, and generative models. The paper discusses the similarities, strengths, and challenges of these models, examines widely used datasets, and reports performance metrics. It also explores promising future research directions in the field. The survey categorizes deep learning-based works into 10 main categories based on their technical contributions. These include fully convolutional networks, convolutional models with graphical models, encoder-decoder models, multi-scale and pyramid network-based models, R-CNN-based models for instance segmentation, dilated convolutional models and DeepLab family, recurrent neural network-based models, attention-based models, generative models and adversarial training, and CNN models with active contour models. The paper provides an overview of around 20 popular image segmentation datasets, grouped into 2D, 2.5D (RGB-D), and 3D images. It also presents a comparative summary of the properties and performance of the reviewed methods on popular benchmarks. The paper discusses the main challenges and future directions for deep learning-based image segmentation, including the need for more efficient models, better handling of context, and improved performance on complex tasks. It also highlights the importance of transfer learning and the use of pre-trained models for segmentation tasks. The survey concludes with a discussion of the current state of the field and the potential for future advancements in deep learning-based image segmentation.