使用 SIFT 功能匹配模板和目标图像之间的点。SIFT 由比例、方向和 128 维局部特征描述符（整数）、f ∈ Z128 组成。您将使用 SIFT 功能在两个图像 I1 和 I2 之间进行匹配。使用模板和目标中的两组描述符，使用比率测试的最近邻查找匹配项。您可以在 MATLAB 中使用 knnsearch 内置函数。函数 [x1， x2] = FindMatch（I1， I2） 输入：两个输入 uint8 格式的灰度图像。输出：x1 和 x2 是指定对应关系的 n × 2 个矩阵。描述：每行 x1 和 x2 分别包含 I1 ad I2 中点对应的 （x， y） 坐标，即 x1（i，:)↔ x2（i，:).（注）您只能使用 VLFeat 进行 SIFT 描述符提取。与比率测试匹配需要自己实现

The purpose of using SIFT features in image matching is to find correspondences between points in two images that are robust to changes in scale, orientation, and illumination. SIFT works by first detecting interest points in an image, which are areas with high variation in intensity. Then, for each interest point, SIFT extracts a 128-dimensional feature vector that describes the local appearance of the region around the point. These feature vectors are then used to match corresponding points in two images, typically using nearest-neighbor matching with a ratio test to ensure that the best match is significantly better than the second-best match. This allows for accurate and robust matching of points between images, even when the images have different scales, orientations, or lighting conditions.