US 12,292,538 B2
Method and system for positioning and correcting visual data by seafloor topographic profiles
Shijuan Yan, Qingdao (CN); Xinyu Shi, Qingdao (CN); Chuanshun Li, Qingdao (CN); Jun Ye, Qingdao (CN); Gang Yang, Qingdao (CN); Zhiwei Zhu, Qingdao (CN); Yue Hao, Qingdao (CN); and Dewen Du, Qingdao (CN)
Assigned to The First Institute of Oceanography, MNR, Qingdao (CN); and Qingdao National Laboratory of Marine Science and Technology Development Center, Qingdao (CN)
Filed by The First Institute of Oceanography, MNR, Qingdao (CN); and Qingdao National Laboratory of Marine Science and Technology Development Center, Qingdao (CN)
Filed on Dec. 25, 2022, as Appl. No. 18/088,599.
Claims priority of application No. 202111664024.3 (CN), filed on Dec. 31, 2021.
Prior Publication US 2023/0213674 A1, Jul. 6, 2023
Int. Cl. G01V 1/38 (2006.01); G06F 18/22 (2023.01)
CPC G01V 1/3817 (2013.01) [G01V 1/3808 (2013.01); G06F 18/22 (2023.01)] 1 Claim
OG exemplary drawing
 
1. A method for positioning and correcting visual data by seafloor topographic profiles, comprising:
step S1: acquiring an original data sequence named a corresponding to a water-depth profile of a target survey line;
step S2: offsetting the water-depth profile of the target survey line equidistantly in a grid layer of a target area to make profiles generated after offsetting traverse the grid layer of the target area, and obtaining offset data sequences named b corresponding to the water-depth profile of the target survey line;
step S3: drawing offset topographic profiles in batches based on offset data of the offset data sequences named b corresponding to the water-depth profile of the target survey line by taking a distance as a horizontal coordinate and an elevation value as a vertical coordinate; wherein ones of the offset topographic profiles whose offset distances each are less than a set offset distance are called seafloor topographic profiles;
step S4: calculating, based on the original data sequence named a and the offset data sequences named b corresponding to the water-depth profile of the target survey line, a profile similarity between the water-depth profile of the target survey line and each of the offset topographic profiles by using a dynamic time warping (DTW) algorithm;
wherein the step S4 comprises:
step S41: calculating, based on the original data sequence named a and each of the offset data sequences named b corresponding to the water-depth profile of the target survey line, a dynamic time warping value by using the DTW algorithm, with the formula as follows:

OG Complex Work Unit Math
where DTW (a, b) represents the dynamic time warping value, dist (a1, b1) represents a first data point in a profile data sequence that is not 0, min represents a function configured to calculate a sum of minimum distances among all other data points except the first data point, m represents a length of the offset data sequence named b, n represents a length of the original data sequence named a, and n and m represent positive integers greater than 2 in the min function;
step S42: calculating an original distance between the original data sequence named a and each of the offset data sequences named b based on the dynamic time warping value;
step S43: calculating a distance similarity between the original data sequence named a and each of the offset data sequences named b based on the original distance between the original data sequence named a and each of the offset data sequences named b;
step S44: taking the original distance corresponding to a largest distance similarity as a shortest distance, and taking a path corresponding to the shortest distance as an optimal path;
step S45: calculating a penalty coefficient based on a number of nodes of the optimal path and lengths of diagonal lines during a calculation of the shortest distance, with the formula as follows:

OG Complex Work Unit Math
where comLeni represents a length of an i-th diagonal line, seqLen represents the number of nodes of the optimal path, and α represents the penalty coefficient;
step S46: calculating the profile similarity between the water-depth profile of the target survey line and each of the offset topographic profiles based on the penalty coefficient and the original distance between the original data sequence named a and each of the offset data sequences named b, with the formula as follows:

OG Complex Work Unit Math
where distance represents the original distance between the original data sequence named a and the offset data sequence named b, α represents the penalty coefficient, and up_similarity represents the profile similarity between the water-depth profile of the target survey line and each of the offset topographic profiles; and
step S5: selecting a geographic location of one of the offset topographic profiles with a largest profile similarity as an actual geographic location of a water-depth profile of a seafloor visual survey line.