Method and system for monitoring analogue gauges using computer vision operations

A system comprising:

• a memory storing instructions;
• one or more Input/Output (I/O) interfaces;
• and
• one or more hardware processors coupled to the memory via the one or more I/O interfaces, wherein the one or more hardware processors are configured by the instructions to:
• receive a real-time feed of an analogue gauge from a set of pre-arranged digital cameras, wherein the analogue gauge is used to measure continuously changing physical quantities;
• pre-process the received real-time feed of an analogue gauge using a low pass filter to reduce noise and sharpness of at least one image of the real-time feed, wherein an adaptive thresholding is performed on the at least one image based on values of pixels around a region with different lighting conditions to obtain edges;
• identify one or more circles of the pre-processed at least one image using a Hough circle transform technique;
• identify one or more edge points of the pre-processed at least one image using a gradient information;
• select at least one of the plurality of circles with a minimum radius to cover a predefined threshold circular region;
• create a radial spokes ring using center of the selected at least one circle, wherein an optimal width of the radial spokes ring is determined to capture the plurality of scales in the predefined threshold circular region;
• determine at least one line of each of a plurality of scales corresponding to the identified one or more circles using a Hough line transform, wherein the plurality of scales is extrapolated towards the center of the identified one or more circles to get a scale region a needle is allowed to traverse;
• analyze the scale region allowed to traverse to determine a minimum and a maximum value of the scale at extreme ends of the circle, wherein the scale is segmented into a predefined parts to increase precision of a meter reading;
• and
• segment the scale region using pixel intensity to check in which segment the needle is present, wherein a pixel mean is calculated for each segment and in the segment with lowest intensity is the one containing the needle.

One or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause:

• receiving, via an input/output (I/O) interface, a real-time feed of an analogue gauge from a set of pre-arranged digital cameras, wherein the analogue gauge is used to measure continuously changing physical quantities;
• pre-processing the received real-time feed of an analogue gauge using a low pass filter to reduce noise and sharpness of at least one image of the real-time feed, wherein an adaptive thresholding is performed on the at least one image based on values of pixels around a region with different lighting conditions to obtain edges;
• identifying one or more circles of the pre-processed at least one image using a Hough circle transform technique;
• identifying one or more edge points of the pre-processed at least one image using a gradient information:
• selecting at least one of the plurality of circles with a minimum radius to cover a predefined threshold circular region;
• creating a radial spokes ring using center of the selected at least one circle, wherein an optimal width of the radial spokes ring is determined to capture the plurality of scales in the predefined threshold circular region;
• determining at least one line of each of a plurality of scales corresponding to the identified one or more circles using a Hough line transform, wherein the plurality of scales is extrapolated towards the center of the identified one or more circles to get a scale region a needle is allowed to traverse;
• analyzing the scale region allowed to traverse to determine a minimum and a maximum value of the scale at extreme ends of the circle, wherein the scale is segmented into a predefined parts to increase precision of a meter reading;
• and
• segmenting the scale region using pixel intensity to check in which segment the needle is present, wherein a pixel mean is calculated for each segment and in the segment with lowest intensity is the one containing the needle.