Systems, methods, and devices for automated meter reading for smart field patrol

A system comprising:

• a transport instrument for patrolling an environment comprising a plurality of measurement devices;
• an image capture device located on the transport instrument and that supports capture of images of the environment;
• a processor that processes an image captured by the image capture device, the image including a first measurement device of the plurality of measurement devices;
• and
• a memory coupled with the processor, wherein the memory stores data that, when executed by the processor, enables the processor to:
• detect a target region included in the image, the target region comprising at least a portion of the first measurement device, wherein detecting the target region comprises:
• providing the image to a machine learning network;
• and
• receiving an output from the machine learning network in response to the machine learning network processing the image based at least in part on a detection model, the output comprising the target region;
• identify the first measurement device from among the plurality of measurement devices based on detecting the target region, wherein identifying the first measurement device from among the plurality of measurement devices includes:
• detecting whether the image comprises an identification tag that identifies the first measurement device, the identification tag being associated with identification information of the first measurement device;
• and
• when the identification tag is not detected in the image, adjusting one or more characteristics of the image to yield a modified image and detecting whether the modified image comprises the identification tag;
• and
• extract measurement information associated with the identified first measurement device based at least in part on detecting the target region.

A system comprising:

• a machine learning network;
• a processor;
• and
• a memory coupled with the processor, wherein the memory stores data that, when executed by the processor, enables the processor to:
• receive an image of an environment comprising a measurement device, wherein the image is captured by an image capture device;
• detect a target region included in the image, the target region comprising at least a portion of the measurement device, wherein detecting the target region comprises:
• providing the image to the machine learning network;
• and
• receiving an output from the machine learning network in response to the machine learning network processing the image based at least in part on a detection model, the output comprising the target region, wherein the detection model comprises an object detection model trained based at least in part on a set of reference images associated with a set of measurement devices, and wherein the set of measurement devices comprises at least the measurement device;
• identify the measurement device from among the set of measurement devices by:
• detecting whether the image comprises a tag that identifies the measurement device;
• when the tag is not detected in the image, adjusting one or more characteristics of the image to yield a modified image and detecting whether the modified image comprises the tag;
• when the tag is detected in the image or the modified image, accessing a database containing a plurality of tags associated to the set of measurement devices to identify the measurement device using the detected tag;
• and
• when the tag is not detected in the image or the modified image, determining a current location of the image capture device within the environment and using the current location of the image capture device to lookup the tag in the database to identify the measurement device;
• and
• extract measurement information from the target region of the image based on the identified measurement device.