This paper introduces a method for extracting distinctive and invariant features from images, which can be used for reliable matching between different views of an object or scene. The features are invariant to image scale and rotation and are robust across a wide range of affine distortions, changes in 3D viewpoint, noise, and illumination changes. The features are highly distinctive, allowing a single feature to be correctly matched with high probability against a large database of features from many images. The paper also describes an approach to using these features for object recognition, which involves matching individual features to a database of known object features, identifying consistent clusters of matches, and verifying the matches through least-squares estimation of pose parameters. This method can robustly identify objects among clutter and occlusion while achieving near real-time performance. The key steps in the process include: 1. **Scale-space extrema detection**: Using a difference-of-Gaussian function to identify potential interest points that are invariant to scale and orientation. 2. **Keypoint localization**: Fitting a detailed model to determine the location and scale of key points. 3. **Orientation assignment**: Assigning one or more orientations to each key point based on local image gradient directions. 4. **Keypoint descriptor**: Computing a representation that allows for significant local shape distortion and change in illumination. The Scale-Invariant Feature Transform (SIFT) method, which transforms image data into scale-invariant coordinates relative to local features, is described in detail. The paper also discusses the trade-offs between efficiency and completeness in feature detection, the accuracy of keypoint localization, and the orientation assignment process. The local image descriptor is designed to be highly distinctive and invariant to changes in illumination and 3D viewpoint. The paper concludes with experimental results demonstrating the effectiveness of the SIFT features in various image matching tasks.This paper introduces a method for extracting distinctive and invariant features from images, which can be used for reliable matching between different views of an object or scene. The features are invariant to image scale and rotation and are robust across a wide range of affine distortions, changes in 3D viewpoint, noise, and illumination changes. The features are highly distinctive, allowing a single feature to be correctly matched with high probability against a large database of features from many images. The paper also describes an approach to using these features for object recognition, which involves matching individual features to a database of known object features, identifying consistent clusters of matches, and verifying the matches through least-squares estimation of pose parameters. This method can robustly identify objects among clutter and occlusion while achieving near real-time performance. The key steps in the process include: 1. **Scale-space extrema detection**: Using a difference-of-Gaussian function to identify potential interest points that are invariant to scale and orientation. 2. **Keypoint localization**: Fitting a detailed model to determine the location and scale of key points. 3. **Orientation assignment**: Assigning one or more orientations to each key point based on local image gradient directions. 4. **Keypoint descriptor**: Computing a representation that allows for significant local shape distortion and change in illumination. The Scale-Invariant Feature Transform (SIFT) method, which transforms image data into scale-invariant coordinates relative to local features, is described in detail. The paper also discusses the trade-offs between efficiency and completeness in feature detection, the accuracy of keypoint localization, and the orientation assignment process. The local image descriptor is designed to be highly distinctive and invariant to changes in illumination and 3D viewpoint. The paper concludes with experimental results demonstrating the effectiveness of the SIFT features in various image matching tasks.