System to determine stationary features by autonomous mobile device

A method performed at least in part by an autonomous mobile device (AMD), the method comprising:

• acquiring, using one or more processors of the AMD, a first pair of images comprising (i) a first image of a physical space acquired at a first time from a first camera of a pair of cameras of the AMD and (ii) a second image of the physical space acquired at the first time from a second camera of the pair of cameras of the AMD, wherein at least a portion of a first field of view of the first camera overlaps at least a portion of a second field of view of the second camera;
• acquiring, using the one or more processors of the AMD, a second pair of images comprising (iii) a third image of the physical space acquired at a second time using the first camera and (iv) a fourth image of the physical space acquired at the second time using the second camera;
• determining, using the one or more processors of the AMD, a first feature that is visible in the first image, the second image, the third image, and the fourth image;
• determining, using the one or more processors of the AMD, a first position of the first feature in the physical space relative to the first camera based on the first pair of images;
• determining, using the one or more processors of the AMD, a second position of the first feature in the physical space relative to the first camera based on the second pair of images;
• acquiring, from the first time to the second time, sensor data from one or more sensors of the AMD;
• determining, using the one or more processors of the AMD, first data indicative of one or more of rotation or translation of the pair of cameras from the first time to the second time based on the sensor data;
• determining, using the one or more processors of the AMD, a third position indicative of a predicted location of the first feature at the second time, based on the first position and the first data;
• determining, using the one or more processors of the AMD, a first error value based on the third position and the second position, wherein the first error value is indicative of a variation between the third position and the second position;
• and
• determining, using the one or more processors of the AMD, second data indicative of whether the first feature was stationary with respect to the physical space between at least the first time and the second time based at least in part on the first error value.

An autonomous mobile device (AMD) comprising:

• a pair of cameras comprising a first camera and a second camera;
• an inertial sensor;
• one or more actuators;
• one or more memories storing first computer-executable instructions;
• and
• one or more processors to execute the first computer-executable instructions to:
• acquire a first pair of images of a physical space comprising (i) a first image of the physical space acquired at a first time from the first camera and (ii) a second image of the physical space acquired at the first time from the second camera;
• acquire a second pair of images of the physical space comprising (iii) a third image of the physical space acquired at a second time from the first camera and (iv) a fourth image of the physical space acquired at the second time from the second camera;
• determine a first feature that is visible in the first image, the second image, the third image, and the fourth image;
• determine, based on the first pair of images, a first position of the first feature in the physical space relative to the first camera;
• determine, based on the second pair of images, a second position of the first feature in the physical space relative to the first camera;
• acquire, from the first time to the second time, inertial data from the inertial sensor;
• determine first data indicative of one or more of rotation or translation of the pair of cameras from the first time to the second time based on the inertial data;
• determine a third position indicative of a predicted location of the first feature at the second time, based on the first position and the first data;
• determine a first error value based on the third position and the second position, wherein the first error value is indicative of a variation between the third position and the second position;
• and
• determine, based at least in part on the first error value, second data indicative of the first feature being stationary with respect to the physical space between at least the first time and the second time.