Privacy protection in mobile robot
A mobile robot is configured for operation in a commercial or industrial setting, such as an office building or retail store. The mobile robot may include cameras for capturing images and videos and include microphones for capturing audio of its surroundings. To improve privacy by preventing confidential information from being transmitted, the mobile robot may detect text in images and modify the images to make the text illegible before transmitting the images. The mobile robot may also detect human voice in audio and modify audio to make the human voice unintelligible before transmitting the audio.
1. A mobile robot comprising:
a motorized base configured to move the mobile robot;
a wireless transceiver configured to communicatively couple to a server;
a housing body storing a camera configured to capture images of an environment surrounding the mobile robot and a microphone configured to capture an audio signal from the environment; and
a controller configured to:
receive images of the environment from the camera;
detect text in the images;
responsive to detecting the text, determine whether to modify the images based in part on a semantic map that stores information describing the environment;
responsive to determining to modify the images, generate a bounding box around the text in the images;
modifying the images within the bounding box to make the text illegible; and
transmit the modified images to the server.
2. The mobile robot of claim 1 , wherein determining whether to modify the images comprises:
determining a location of the mobile robot;
querying the semantic map associated with the environment to determine a privacy setting associated with the location; and
determining to modify the images based on the privacy setting associated with the location.
3. The mobile robot of claim 1 , wherein determining whether to modify the images comprises:
performing image recognition to determine an object type associated with an object including the text in the images;
querying the semantic map associated with the environment to determine a privacy setting associated with the object type; and
determining to modify a region of the images including the object based the privacy setting associated with the object type.
4. The mobile robot of claim 1 , wherein detecting text further comprises:
for each of a plurality of regions within the image:
identifying a number of edges in the region;
comparing the number of edges to a predetermined threshold; and
responsive to the number of edges exceeding the predetermined threshold, determining that there is text in the region.
5. The mobile robot of claim 1 , wherein the controller is further configured to:
detect a face of a person in the images;
apply a bounding box around the face in the images;
modify the images within the bounding box to make the face unidentifiable from the images; and
transmit the modified images to the server.
6. The mobile robot of claim 1 , wherein the controller is further configured to:
generate a hash value based on the detected face; and
transmit the hash value to the server in association with the modified images.
7. The mobile robot of claim 1 , wherein the controller is further configured to:
receive the audio signal of the environment from the microphone;
detect human voice in the audio signal;
responsive to detecting the human voice, modify a portion of the audio signal within a frequency range of human voice to be unintelligible; and
transmit the modified audio signal to the server.
8. A method comprising:
capturing, by a mobile robot, images of an environment surrounding the mobile robot;
detecting, by the mobile robot, text in the images;
responsive to detecting the text, determining, by the mobile robot, whether to modify the images based in part on a semantic map that stores information describing the environment;
responsive to determining to modify the images, generating, by the mobile robot, a bounding box around the text in the images;
modifying, by the mobile robot, the images within the bounding box to make the text illegible, and
transmitting, by the mobile robot, the modified images to a server.
9. The method of claim 8 , wherein determining whether to modify the images comprises:
determining a location of the mobile robot;
querying the semantic map associated with the environment to determine a privacy setting associated with the location; and
determining to modify the images based on the privacy settings associated with the location.
10. The method of claim 8 , wherein determining whether to modify the images comprises:
performing image recognition to determine an object type associated with an object including the text in the images;
querying the semantic map associated with the environment to determine a privacy setting associated with the object type; and
determining to modify a region of the images including the object based the privacy setting associated with the object type.
11. The method of claim 8 , wherein detecting text, further comprises:
for each of a plurality of regions within the image:
identifying a number of edges in the region;
comparing the number of edges to a predetermined threshold; and
responsive to the number of edges exceeding the predetermined threshold, determining that there is text in the region.
12. The method of claim 8 , further comprising:
detecting a face of a person in the images;
applying a bounding box around the face in the images;
modifying the images within the bounding box to make the face unidentifiable from the images; and
transmitting the modified images to the server.
13. The method of claim 12 , further comprising:
generate a hash value based on the detected face; and
transmit the hash value to the server in association with the modified images.
14. The method of claim 8 , further comprising:
receiving an audio signal of the environment surrounding the mobile robot;
detecting human voice in the audio signal;
responsive to detecting the human voice, modifying a portion of the audio signal within a frequency range of human voice to be unintelligible; and
transmitting the modified audio signal to the server.
15. A non-transitory computer readable storage medium storing computer program instructions, the instructions when executed by a processor causing the processor to:
capture images of an environment surrounding a mobile robot;
detect text in the images;
responsive to detecting the text, determine whether to modify the images based in part on a semantic map that stores information describing the environment;
responsive to determining to modify the images, generate a bounding box around the text in the images;
modify the images within the bounding box to make the text illegible, and
transmit the modified images to a server.
16. The non-transitory computer-readable storage medium of claim 15 , wherein determining whether to modify the images comprises:
determining a location of the mobile robot;
querying the semantic map associated with the environment to determine a privacy setting associated with the location; and
determining to modify the images based on the privacy settings associated with the location.
17. The non-transitory computer-readable storage medium of claim 15 , wherein determining whether to modify the images comprises:
performing image recognition to determine an object type associated with an object including the text in the images;
querying the semantic map associated with the environment to determine a privacy setting associated with the object type; and
determining to modify a region of the images including the object based the privacy setting associated with the object type.
18. The non-transitory computer-readable storage medium of claim 15 , wherein detecting text, further comprises:
for each of a plurality of regions within the image:
identifying a number of edges in the region;
comparing the number of edges to a predetermined threshold; and
responsive to the number of edges exceeding the predetermined threshold, determining that there is text in the region.
19. The non-transitory computer-readable storage medium of claim 15 , wherein the processors are further configured to:
detect a face of a person in the images;
apply a bounding box around the face in the images;
modify the images within the bounding box to make the face unidentifiable from the images; and
transmit the modified images to the server.
20. The non-transitory computer-readable storage medium of claim 16 , wherein the processors are further configured to:
determine an identity of the person;
generate a hash value associated with the identity; and
transmit the hash value to the server in association with the modified images.