Construction site productivity capture using computer vision

An image capturing device for capturing objects in action at an earthmoving site, the image capturing device comprising:

• a camera for capturing images of a region of the earthmoving site;
• communication circuitry operable to send data to a remote server;
• memory for storing at least a portion of the captured images of the region of the earthmoving site, the memory comprising an image classifier module or an object detection module;
• and
• a processor coupled to the memory, the camera, and the communication circuitry, the processor capable of executing instructions in the memory to:
• capture an image of the region of the earthmoving site including an object, the image being recorded as captured digital data;
• identify a classification of the object using a trained algorithm of the image classifier module or the object detection module existing in the memory;
• send, in response to identifying the classification of the object, the classification of the object to the remote server through a network;
• determine, by the object detection module, a pixel location and a boundary of the object within a field of view of the image capturing device based on positions of pixels of the object in the image;
• identify a triggering event in the image based on the classification of the object, the pixel location, and the boundary of the object
• send, in response to identifying the triggering event, a set of images including the image as digital data to the remote server, the set of images representing a movement of the object at the earthmoving site;
• determine, in response to identifying the triggering event, an activity being performed by the object based on an analysis of the digital data associated with the set of images, the classification of the object, and the movement of the object at the earthmoving site;
• and
• output a report to a user based on the activity performed by the object at the earthmoving site.

A system for capturing an object in action at an earthmoving site, the system comprising:

• a first image capturing device positioned within the earthmoving site and operable to capture first images of a first region of the earthmoving site;
• and
• a second image capturing device positioned within the earthmoving site and configured to capture second images of a second region of the earthmoving site, wherein the first region overlaps with at least a portion of the second region, and wherein the first image capturing device and the second image capturing device are communicatively coupled to a remote server over a network,
• wherein the first image capturing device comprises:
• a camera for capturing the first images of the first region;
• communication circuitry operable to send data to the remote server;
• memory for storing at least a portion of the captured first images of the first region of the earthmoving site, the memory comprising an image classifier module and an object detection module;
• and
• a processor coupled to the memory, the camera, and the communication circuitry, the processor capable of executing instructions in the memory to:
• capture an image of a region of the earthmoving site including an object, the image being recorded as captured digital data;
• identify a classification of the object using a trained algorithm existing in the memory of the first image capturing device;
• send, in response to identifying the classification of the object, the classification of the object to the remote server through the network;
• determine a pixel location and a boundary of the object within a field of view of the first image capturing device based on positions of pixels of the object in the image;
• identify a triggering event in the image based on the classification of the object, the pixel location, and the boundary of the object
• send, in response to identifying the triggering event, a set of images including the image as digital data to the remote server, the set of images representing a movement of the object at the earthmoving site;
• and
• determine, in response to identifying the triggering event, an activity being performed by the object based on an analysis of the digital data associated with the set of images, the classification of the object, and the movement of the object at the earthmoving site,
• wherein a report based on the activity performed by the object at the earthmoving site is outputted to a user.