This paper describes the split measure algorithm about tracks data. In accordance with the needs of the project, the algorithm can be used to measure the two parts, the three parts and the five parts of the data and then combine them together. The algorithm is based on the overlap area of each measurement. And the best stitching position can be obtained by using the least square method. Error analysis of two parts, three parts and five parts measure respectively is carried out. For several measure requirements, Error analysis shows that the theoretical maximum error of the algorithm is less than the actual requirements of the project. Finally, we selected 300 data from a sine wave, and using these 300 data to simulate the 3 meters long track. Then to simulate five measurements, and every two measurements overlap 0.5 meters. Experimental data analysis shows that the algorithm can indeed meet the requirements of practical application.

Abstract-Corner detection algorithms are widely used in image recognition, image matching and edge detection. Accurately finding a corner in a image can greatly improves the algorithms accuracy of the above research areas. So many scientists and researchers have begun to the hot research field and have made much remarkable achievements. The paper introduced the theory of classic corner detection algorithm, including Harris algorithm, surf and sift algorithm, ORB algorithm, Susan algorithm and so on, And compared the difference of classic algorithms. Then, the recently outstanding algorithm and its application are introduced. Finally, the paper summarizes the research trends about corner detection algorithms.