# US 12,112,548 B1 · Detection of camera with impaired view > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,112,548 B1 - Document ID: US12112548B1 - Status: Granted - Kind: B1 - Application number: 18/618,182 - Filed: 2024-03-27 - Granted: 2024-10-08 - Earliest claimed: 2024-03-27 - Assignees: Samsara Inc. - Applicants: Samsara Inc. - Inventors: Shaurye Agarwal; Akshay Raj Dhamija; Howard Yu; Margaret Irene Finch; Jing Wang; Rohit Annigeri; Sharan Srinivasan; Yan Wang; Nathan Hurst - Web page: https://patentgrep.com/patent/US12112548B1 ## CPC - G06V 20/56: exterior to a vehicle by using sensors mounted on the vehicle - G06T 3/60: Rotation of whole images or parts thereof - G06V 10/44: Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components - G06V 10/764: using classification, e.g. of video objects - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract Methods, systems, and programs are presented for detecting impaired views in monitoring cameras. One method includes training a rotation classifier with unsupervised learning utilizing a first set of images. The rotation classifier is configured to receive an input image and generate a rotation feature embedding for the input image. In addition, the method includes training an impairment classifier with supervised learning utilizing a second set of images, impairment labels for each of the second set of images, and the rotation feature embedding, generated by the rotation classifier, for each of the second set of images. The method further includes accessing a vehicle image captured by a camera on a vehicle, and providing the vehicle image to the impairment classifier as input, and the impairment classifier outputs a camera impairment from a set of camera impairment categories. Further, the vehicle image and the camera impairment are presented on a user interface. ## Claims ### Claim 1 A computer-implemented method comprising: training a rotation classifier with unsupervised learning utilizing a first plurality of images, the rotation classifier configured to receive an input image and generate a rotation feature embedding for the input image; training an impairment classifier with supervised learning utilizing a second plurality of images, impairment labels for each image of the second plurality of images, and the rotation feature embedding, generated by the rotation classifier, for each of the second plurality of images; accessing a vehicle image captured by a camera on a vehicle; providing the vehicle image to the impairment classifier as input, the impairment classifier providing an output of a camera impairment from a set of camera impairment categories; and causing presentation on a user interface (UI) of the vehicle image and the camera impairment. ### Claim 2 (depends on 1) The method as recited in claim 1, further comprising: determining, at a behavioral monitoring system (BMS), that the camera impairment is fixable based on a rotation of the vehicle image; and rotating, at the BMS, images received from the vehicle for behavioral monitoring of the vehicle. ### Claim 3 (depends on 1) The method as recited in claim 1, further comprising: determining, at a behavioral monitoring system (BMS), that the camera impairment is fixable based on a rotation of the vehicle image; and sending, from the BMS to a network gateway device (NGD) at the vehicle, a request to rotate the images captured by the camera in the vehicle. ### Claim 4 (depends on 1) The method as recited in claim 1, wherein the camera impairment corresponds to a camera that is unable to provide clear and accurate images required for vehicle behavioral monitoring. ### Claim 5 (depends on 1) The method as recited in claim 1, wherein the training of the impairment classifier further includes information about daytime or nighttime associated with each image from the second plurality of images, wherein the impairment classifier further receives as input the information about daytime or nighttime associated with the vehicle image. ### Claim 6 (depends on 1) The method as recited in claim 1, further comprising: monitoring a fleet of vehicles to determine camera impairment for cameras installed in the fleet of vehicles, the monitoring comprising checking for camera impairment based on images captured at a beginning or an end of trips of the vehicles in the fleet of vehicles. ### Claim 7 (depends on 1) The method as recited in claim 1, wherein the camera impairment categories comprise two or more of normal; hard obstruction 25-50%; hard obstruction 50-75%; hard obstruction 75-100%; misalignment titled up; misalignment tilted down; misalignment upside down; misalignment rotated; misalignment reversed; smudges; external vehicle artifacts; not attached to vehicle; normal with snow or fog; and normal inside building, tunnel, bridges, parking lot, or walls. ### Claim 8 (depends on 1) The method as recited in claim 1, wherein the camera impairment categories comprise two or more of day normal, night normal, misaligned, center artifact, partial obstruction, smudged, hard obstruction, and weather. ### Claim 9 (depends on 1) The method as recited in claim 1, further comprising: causing an alert for a driver of the vehicle in response to the camera impairment. ### Claim 10 (depends on 1) The method as recited in claim 1, further comprising: generating a report of camera impairment events for a fleet of vehicles; and causing presentation in the UI of the report. ### Claim 11 A system comprising: a memory comprising instructions; and one or more computer processors, wherein the instructions, when executed by the one or more computer processors, cause the system to perform operations comprising: training a rotation classifier with unsupervised learning utilizing a first plurality of images, the rotation classifier configured to receive an input image and generate a rotation feature embedding for the input image; training an impairment classifier with supervised learning utilizing a second plurality of images, impairment labels for each image of the second plurality of images, and the rotation feature embedding, generated by the rotation classifier, for each of the second plurality of images; accessing a vehicle image captured by a camera on a vehicle; providing the vehicle image to the impairment classifier as input, the impairment classifier providing an output of a camera impairment from a set of camera impairment categories; and causing presentation on a user interface (UI) of the vehicle image and the camera impairment. ### Claim 12 (depends on 11) The system as recited in claim 11, wherein the instructions further cause the one or more computer processors to perform operations comprising: determining, at a behavioral monitoring system (BMS), that the camera impairment is fixable based on a rotation of the vehicle image; and rotating, at the BMS, images received from the vehicle for behavioral monitoring of the vehicle. ### Claim 13 (depends on 11) The system as recited in claim 11, wherein the instructions further cause the one or more computer processors to perform operations comprising: determining, at a behavioral monitoring system (BMS), that the camera impairment is fixable based on a rotation of the vehicle image; and sending, from the BMS to a network gateway device (NGD) at the vehicle, a request to rotate the images captured by the camera in the vehicle. ### Claim 14 (depends on 11) The system as recited in claim 11, wherein the camera impairment corresponds to a camera that is unable to provide clear and accurate images required for vehicle behavioral monitoring. ### Claim 15 (depends on 11) The system as recited in claim 11, wherein the training of the impairment classifier further includes information about daytime or nighttime associated with each image from the second plurality of images, wherein the impairment classifier further receives as input the information about daytime or nighttime associated with the vehicle image. ### Claim 16 A non-transitory machine-readable storage medium including instructions that, when executed by a machine, cause the machine to perform operations comprising: training a rotation classifier with unsupervised learning utilizing a first plurality of images, the rotation classifier configured to receive an input image and generate a rotation feature embedding for the input image; training an impairment classifier with supervised learning utilizing a second plurality of images, impairment labels for each image of the second plurality of images, and the rotation feature embedding, generated by the rotation classifier, for each of the second plurality of images; accessing a vehicle image captured by a camera on a vehicle; providing the vehicle image to the impairment classifier as input, the impairment classifier providing an output of a camera impairment from a set of camera impairment categories; and causing presentation on a user interface (UI) of the vehicle image and the camera impairment. ### Claim 17 (depends on 16) The non-transitory machine-readable storage medium as recited in claim 16, wherein the machine further performs operations comprising: determining, at a behavioral monitoring system (BMS), that the camera impairment is fixable based on a rotation of the vehicle image; and rotating, at the BMS, images received from the vehicle for behavioral monitoring of the vehicle. ### Claim 18 (depends on 16) The non-transitory machine-readable storage medium as recited in claim 16, wherein the machine further performs operations comprising: determining, at a behavioral monitoring system (BMS), that the camera impairment is fixable based on a rotation of the vehicle image; and sending, from the BMS to a network gateway device (NGD) at the vehicle, a request to rotate the images captured by the camera in the vehicle. ### Claim 19 (depends on 16) The non-transitory machine-readable storage medium as recited in claim 16, wherein the camera impairment corresponds to a camera that is unable to provide clear and accurate images required for vehicle behavioral monitoring. ### Claim 20 (depends on 16) The non-transitory machine-readable storage medium as recited in claim 16, wherein the training of the impairment classifier further includes information about daytime or nighttime associated with each image from the second plurality of images, wherein the impairment classifier further receives as input the information about daytime or nighttime associated with the vehicle image.