Systems and methods for automatically extracting objects from images
A method for identifying and extracting objects from a video frame includes accessing a video frame of a video. The method further includes analyzing the video frame to generate a plurality of masks. Each mask includes neighboring pixels that are determined to be related. The method further includes extracting a plurality of objects from the video frame based on the generated plurality of masks. The method further includes recursively extracting a plurality of nested objects from the video frame based on the generated plurality of masks. The method further includes creating a plurality of object images by combining the plurality of masks of the plurality of objects with image data of the video frame and combining the plurality of masks of the plurality of nested objects with the image data of the video frame. The method further includes displaying the plurality of object images in a graphical user interface.
1 . A system comprising:
a camera;
one or more memory units; and
one or more computer processors communicatively coupled to the one or more memory units and configured to perform operations comprising:
access a video frame of a video generated by the camera when viewing a physical environment;
analyze the video frame of the video in order to generate a plurality of masks, each mask comprising a set of neighboring pixels that are determined to be related;
extract a plurality of objects from the video frame based on the generated plurality of masks;
recursively extract a plurality of nested objects from the video frame based on the generated plurality of masks, each particular nested object being related to a particular one of the plurality of objects;
create a plurality of object images by:
combining the plurality of masks of the plurality of objects with image data of the video frame; and
combining the plurality of masks of the plurality of nested objects with the image data of the video frame; and
display one or more of the plurality of object images in a graphical user interface.
2 . The system of claim 1 , wherein displaying the one or more of the plurality of object images in the graphical user interface is based on a user input, the user input indicating a mask level of the plurality of object images.
3 . The system of claim 1 , wherein analyzing the video frame of the video in order to generate the plurality of masks comprises utilizing an image segmentation algorithm.
4 . The system of claim 1 , wherein each mask comprises a binary bitmap image.
5 . The system of claim 1 , wherein extracting the plurality of objects from the video frame based on the generated plurality of masks comprises:
identifying each particular mask of the plurality of masks as an object of the plurality of objects; and
filtering the plurality of objects to remove any objects determined to be background objects.
6 . The system of claim 1 , the operations further comprising displaying a bounding box around each of the plurality of objects and each of the plurality of nested objects on the video frame.
7 . The system of claim 1 , wherein each of the plurality of object images comprises a cropped image of a corresponding object or nested object on a solid-color background.
8 . A method by a computing system for identifying and extracting objects from images, the method comprising:
accessing a video frame of a video generated by a camera when viewing a physical environment;
analyzing the video frame of the video in order to generate a plurality of masks, each mask comprising a set of neighboring pixels that are determined to be related;
extracting a plurality of objects from the video frame based on the generated plurality of masks;
recursively extracting a plurality of nested objects from the video frame based on the generated plurality of masks, each particular nested object being related to a particular one of the plurality of objects;
creating a plurality of object images by:
combining the plurality of masks of the plurality of objects with image data of the video frame; and
combining the plurality of masks of the plurality of nested objects with the image data of the video frame; and
displaying one or more of the plurality of object images in a graphical user interface.
9 . The method of claim 8 , wherein displaying the one or more of the plurality of object images in the graphical user interface is based on a user input, the user input indicating a mask level of the plurality of object images.
10 . The method of claim 8 , wherein analyzing the video frame of the video in order to generate the plurality of masks comprises utilizing an image segmentation algorithm.
11 . The method of claim 8 , wherein each mask comprises a binary bitmap image.
12 . The method of claim 8 , wherein extracting the plurality of objects from the video frame based on the generated plurality of masks comprises:
identifying each particular mask of the plurality of masks as an object of the plurality of objects; and
filtering the plurality of objects to remove any objects determined to be background objects.
13 . The method of claim 8 , further comprising displaying a bounding box around each of the plurality of objects and each of the plurality of nested objects on the video frame.
14 . The method of claim 8 , wherein each of the plurality of object images comprises a cropped image of a corresponding object or nested object on a solid-color background.
15 . One or more computer-readable non-transitory storage media embodying instructions that, when executed by a processor, cause the processor to perform operations comprising:
access a video frame of a video generated by a camera when viewing a physical environment;
analyze the video frame of the video in order to generate a plurality of masks, each mask comprising a set of neighboring pixels that are determined to be related;
extract a plurality of objects from the video frame based on the generated plurality of masks;
recursively extract a plurality of nested objects from the video frame based on the generated plurality of masks, each particular nested object being related to a particular one of the plurality of objects;
create a plurality of object images by:
combining the plurality of masks of the plurality of objects with image data of the video frame; and
combining the plurality of masks of the plurality of nested objects with the image data of the video frame; and
display one or more of the plurality of object images in a graphical user interface.
16 . The one or more computer-readable non-transitory storage media of claim 15 , wherein displaying the one or more of the plurality of object images in the graphical user interface is based on a user input, the user input indicating a mask level of the plurality of object images.
17 . The one or more computer-readable non-transitory storage media of claim 15 , wherein analyzing the video frame of the video in order to generate the plurality of masks comprises utilizing an image segmentation algorithm.
18 . The one or more computer-readable non-transitory storage media of claim 15 , wherein each mask comprises a binary bitmap image.
19 . The one or more computer-readable non-transitory storage media of claim 15 , wherein extracting the plurality of objects from the video frame based on the generated plurality of masks comprises:
identifying each particular mask of the plurality of masks as an object of the plurality of objects; and
filtering the plurality of objects to remove any objects determined to be background objects.
20 . The one or more computer-readable non-transitory storage media of claim 15 , the operations further comprising displaying a bounding box around each of the plurality of objects and each of the plurality of nested objects on the video frame.