US 11,656,328 B2
Validating object detection hardware and algorithms
Aviram Shmueli, Modi'in (IL); Alon Ascher, Tel Aviv (IL); Ben Alfi, Tel Aviv (IL); and Shmuel Ur, Shorashim (IL)
Assigned to Blue White Robotics Ltd, Tel Aviv (IL)
Filed by Blue White Robotics Ltd, Tel Aviv (IL)
Filed on Jun. 2, 2020, as Appl. No. 16/889,966.
Prior Publication US 2021/0374438 A1, Dec. 2, 2021
Int. Cl. G01S 13/86 (2006.01); G01S 7/40 (2006.01); G01S 13/931 (2020.01); G01S 17/931 (2020.01); G06V 20/58 (2022.01); G06F 18/22 (2023.01); G06V 10/776 (2022.01)
CPC G01S 7/4004 (2013.01) [G01S 13/867 (2013.01); G01S 13/931 (2013.01); G01S 17/931 (2020.01); G06F 18/22 (2023.01); G06V 10/776 (2022.01); G06V 20/58 (2022.01); G01S 2013/932 (2020.01)] 18 Claims
OG exemplary drawing
 
1. A method of automatically evaluating detection accuracy of dynamic objects by equipment under test, comprising:
using at least one processor for:
receiving a first record generated by an evaluated equipment under test which is deployed in a vehicle, the first record comprising a plurality of attributes of each dynamic object detected by the evaluated equipment in an environment of the vehicle;
receiving a second record generated by a validated reference equipment deployed in the vehicle, the second record comprising a plurality of attributes of each dynamic object detected in the environment by the reference equipment;
identifying at least one dynamic object appearing in both said first record and said second record by detecting when said at least one dynamic object occupies a same space at a same time in both said first record and said second record, by correlating between the at least one dynamic object detected by the evaluated equipment and a respective dynamic object detected by the reference equipment according to a match between at least one spatial attribute and at least one temporal attribute of the plurality of attributes of the respective dynamic object in the first record and in the second record;
analyzing at least some of the plurality of attributes of the respective dynamic object in the first record compared to respective attributes of the respective dynamic object in the second record;
analyzing sensory data captured by the evaluated equipment compared to corresponding sensory data captured by the validated reference equipment to evaluate performance of the evaluated equipment;
outputting an indication of at least one difference between the first record and the second record identified based on the analysis; and
outputting evaluation results of the analysis of the performance of the evaluated equipment.