Measuring people-flow through doorways using easy-to-install IR array sensors

A sensor system for measuring people flow through a doorway, the sensor system comprising:

• an infrared array having a plurality of infrared sensors arranged in as a grid and oriented to view the doorway;
• and
• a processor operably connected to the infrared array, the processor being configured to:
• read a sequence of sensor data frames from the infrared array, each sensor data frame having a grid of pixels, each pixel having a temperature value;
• determine a background temperature frame as a pixel-wise average of a predetermined number of initial sensor data frames from the sequence of sensor data frames, the predetermined number of initial sensor data frames including at least two initial sensor data frames;
• determine, for each sensor data frame in the sequence of sensor data frames after the predetermined number of initial sensor data frames, whether each respective sensor data frame includes at least one person using at least one filtering technique that identifies sensor data frames that do not include at least one person;
• identify, for each sensor data frame in the sequence of sensor data frames that includes at least one person, at least one group of adjacent pixels in the respective sensor data frame corresponding to the at least one person by (i) comparing each pixel in the respective sensor data frame with a first threshold, in each case the first threshold being equal to a value of a corresponding pixel in the background temperature frame plus a predetermined temperature, (ii) generating an intermediate data frame based on the comparison, pixels of the intermediate data frame being set to a first value in response to corresponding pixels in the respective sensor data frame having a value greater than the corresponding first threshold and set to a second value in response to corresponding pixels in the respective sensor data frame having a value less than the corresponding first threshold, (iii) identifying at least one group of adjacent pixels in the intermediate data frame having the first value, and (iv) in response to the at least one group of adjacent pixels in the intermediate data frame having a size less than a first predetermined percentage of a total size of the intermediate data frame, identifying the at least one group of adjacent pixels in the respective sensor data frame as pixels in the respective sensor data frame corresponding to the at least one group of adjacent pixels in the intermediate data frame;
• determine, for each sensor data frame in the sequence of sensor data frames that includes at least one person, at least one location in the respective sensor data frame corresponding to the at least one person;
• and
• determine, for each sensor data frame in the sequence of sensor data frames determined that includes at least one person, whether the at least one person in the respective sensor data frame is a same person as a person in a previous sensor data frame in the sequence of sensor data frames.

A method of measuring people flow through a doorway, the method comprising:

• reading, with a processor, a sequence of sensor data frames from an infrared array that views the doorway, each sensor data frame having a grid of pixels, each pixel having a temperature value;
• determining, with the processor, a background temperature frame as a pixel-wise average of a predetermined number of initial sensor data frames from the sequence of sensor data frames, the predetermined number of initial sensor data frames including at least two initial sensor data frames;
• determining, with the processor, for each sensor data frame in the sequence of sensor data frames after the predetermined number of initial sensor data frames, whether each respective sensor data frame includes at least one person using at least one filtering technique that identifies sensor data frames that do not include at least one person;
• identifying, with the processor, for each sensor data frame in the sequence of sensor data frames that includes at least one person, at least one group of adjacent pixels in the respective sensor data frame corresponding to the at least one person;
• determining, with the processor, for each sensor data frame in the sequence of sensor data frames that includes at least one person, at least one location in the respective sensor data frame corresponding to the at least one person;
• and
• determining, with the processor, for each sensor data frame in the sequence of sensor data frames determined that includes at least one person, whether the at least one person in the respective sensor data frame is a same person as a person in a previous sensor data frame in the sequence of sensor data frames,
• wherein at least one of:
• the determining at least one location in each respective sensor data frame corresponding to the at least one person further comprises (i) determining sums of values of pixels in one of each column and each row in the respective sensor data frame, (ii) determining at least one maximum of the sums, and (iii) determining the at least one location as one of at least one column and at least one row in the respective sensor data frame corresponding the at least one maximum, wherein if the at least one person includes only one person, then the at least one maximum is an absolute maximum of the sums, wherein if the at least one person includes at least two people, then the at least one maximum includes at least two largest local maximums of the sums;
• and
• the determining whether the at least one person in each respective sensor data frame is a same person as a person in a previous sensor data frame in the sequence of sensor data frames further comprises (I) determining at least one first difference between the at least one location of the at least one person in the respective sensor data frame and a location of the person in the previous sensor data frame, (II) determining that the at least one person in the respective sensor data frame is not the same person as the person in the previous sensor data frame in response to the at least one first difference being greater than a first predetermined threshold, and (III) in response to the at least one first difference being less than the first predetermined threshold:
• determining at least one first average value of pixels in the at least one group of pixels corresponding to the at least one person in the respective sensor data frame, determining a second average value of pixels in a group of pixels corresponding to the person in the previous sensor data frame, determining at least one second difference between the at least one first average value and the second average value, determining that the at least one person in the respective sensor data frame is the same person as the person in the previous sensor data frame in response to the at least one second difference being less than a second predetermined threshold, and determining that the at least one person in the respective sensor data frame is not the same person as the person in the previous sensor data frame in response to the at least one second difference being greater than the second predetermined threshold.