This paper provides a comprehensive review of recent progress in object detection in optical remote sensing images (RSIs). It covers a wide range of applications, including environmental monitoring, geological hazard detection, land-use/land-cover (LULC) mapping, geographic information system (GIS) update, precision agriculture, and urban planning. The paper discusses the challenges faced in object detection, such as viewpoint variation, occlusion, background clutter, illumination, and shadow, and highlights the advancements in remote sensing technology that have enabled the detection of more detailed spatial and textural information. The review focuses on generic object categories, including roads, buildings, trees, vehicles, ships, airports, and urban areas, rather than specific classes like buildings or roads. It covers 270 publications and examines five main categories of methods: template matching-based, knowledge-based, object-based image analysis (OBIA)-based, machine learning-based, and discusses five publicly available datasets and three standard evaluation metrics. Key topics include: 1. **Template Matching-Based Methods**: These methods use pre-defined templates to detect objects, categorized into rigid and deformable templates. 2. **Knowledge-Based Methods**: These methods leverage geometric and context information to detect objects. 3. **OBIA-Based Methods**: These methods involve image segmentation and object classification to detect objects. 4. **Machine Learning-Based Methods**: These methods use feature extraction, feature fusion, dimension reduction, and classifier training to detect objects. 5. **Datasets and Evaluation Metrics**: The paper reviews five datasets and three standard evaluation metrics for object detection. The paper also discusses open problems and challenges in current studies and proposes two promising research directions: deep learning-based feature representation and weakly supervised learning-based geospatial object detection. The aim is to provide researchers with a better understanding of the field and to advance the development of more effective object detection frameworks.This paper provides a comprehensive review of recent progress in object detection in optical remote sensing images (RSIs). It covers a wide range of applications, including environmental monitoring, geological hazard detection, land-use/land-cover (LULC) mapping, geographic information system (GIS) update, precision agriculture, and urban planning. The paper discusses the challenges faced in object detection, such as viewpoint variation, occlusion, background clutter, illumination, and shadow, and highlights the advancements in remote sensing technology that have enabled the detection of more detailed spatial and textural information. The review focuses on generic object categories, including roads, buildings, trees, vehicles, ships, airports, and urban areas, rather than specific classes like buildings or roads. It covers 270 publications and examines five main categories of methods: template matching-based, knowledge-based, object-based image analysis (OBIA)-based, machine learning-based, and discusses five publicly available datasets and three standard evaluation metrics. Key topics include: 1. **Template Matching-Based Methods**: These methods use pre-defined templates to detect objects, categorized into rigid and deformable templates. 2. **Knowledge-Based Methods**: These methods leverage geometric and context information to detect objects. 3. **OBIA-Based Methods**: These methods involve image segmentation and object classification to detect objects. 4. **Machine Learning-Based Methods**: These methods use feature extraction, feature fusion, dimension reduction, and classifier training to detect objects. 5. **Datasets and Evaluation Metrics**: The paper reviews five datasets and three standard evaluation metrics for object detection. The paper also discusses open problems and challenges in current studies and proposes two promising research directions: deep learning-based feature representation and weakly supervised learning-based geospatial object detection. The aim is to provide researchers with a better understanding of the field and to advance the development of more effective object detection frameworks.