For SF6 pressure pointer meter with complex dial, internal mechanical components and short pointer, there is still a lack of a method of meter location identification with low computational cost and robustness and accuracy, which is suitable for poor environment. To solve this problem, this paper proposes a positioning and recognition method of pointer meters based on binary robust invariant scalable keypoints (BRISK) and fast retina keypoint (FREAK) feature detection technology and line segment detector (LSD) for accurate reading. Firstly, the BRISK algorithm feature detector and FREAK algorithm feature descriptor are introduced to detect the dial area of the meter combined with K-nearest neighbor (KNN) matching algorithm, and the dial is automatically corrected. Then, on the basis of comprehensive application of image filtering de-noising, binarization, edge detection and other image pretreatment methods, Hough transform method is used to determine the center of the instrument, LSD algorithm combined with simple constraints to complete pointer line detection. Finally, the angle method is used for reading. The positioning results of a large number of SF6 pressure pointer meter images show that the positioning method in this paper still has good adaptability under harsh conditions such as strong light intensity, large dial tilt degree and blurred dial image, and has high accuracy and practicability for the recognition of SF6 pressure pointer meter.