NIR motion detection system and method
A motion sensor for detection motion of humans is provided. The motion sensor contains a near infrared (NIR) low resolution image sensor that captures image frames in the near infrared spectrum and a sensor that detects the amount of visible light. In addition, a processor is connected to the visible light sensor and the NIR motion sensor. The processor is configured to receive the amount of visible light from the visible light sensor and the images from the NIR low resolution image sensor. The processor is further configured to compare the image frames to detect motion; the sensitivity of the detection of motion is determined by the amount of visible light detected by the visible light sensor. The output has two or modes based on the detection of motion by the processor.
1. A motion sensor comprising:
a NIR sensor to capture a first image and a second image in a near infrared spectrum, wherein the second image is captured after the first image by a period of time;
a processor to generate a subtracted image from the first image and the second image, wherein the subtracted image is to highlight a change over the period of time; and
a classifier to classify the change via a machine learning process, wherein the classifier is trained to classify the change, via supervised training, as representing one of a state of vacancy, a major motion, and a minor motion, and wherein the classifier determines a confidence value associated with a classification; wherein the confidence value is used to control a space lighting system.
2. The motion sensor of claim 1 , further comprising a visible light sensor to measure an intensity of total visible light, wherein the intensity of total visible light is to be used by the classifier.
3. The motion sensor of claim 1 , further comprising a source of light in the near infrared spectrum to provide light to be reflected to the NIR sensor.
4. A method comprising:
capturing a first image in a near infrared spectrum with a NIR sensor;
capturing a second image in the near infrared spectrum with the NIR sensor, wherein the second image is captured after the first image by a period of time;
generating a subtracted image from the first image and the second image, wherein the subtracted image is to highlight a change over the period of time;
training a classifier to classify the change via supervised training, wherein training the classifier comprises training the classifier to classify the change as representing one of a state of vacancy, a major motion, and a minor motion;
classifying the change with a machine learning process;
determining a confidence value associated with a classification; and
controlling a space lighting system based on the confidence value.
5. The method of claim 4 , further comprising measuring an intensity of total visible light, wherein the intensity of total visible light is to be used to classify the change.
6. The method of claim 5 , wherein determining the confidence value is based on the intensity of total visible light.
7. The method of claim 4 , further comprising providing light from a source of light in the near infrared spectrum, wherein the light is to be reflected off a surface to the NIR sensor.
8. An apparatus comprising:
an space lighting system;
a NIR sensor to capture data, wherein the data includes a plurality of images in a near infrared spectrum;
a processor to generate a subtracted image from a first image and a second image, wherein the first image and the second image are successive images from the plurality of images, and wherein the subtracted image is to highlight a change between the first image and the second image; and
a classifier to classify a change between first image and the second image, wherein a classification of the change is used to control the space lighting system.
9. The apparatus of claim 8 , wherein the classifier is trained via a supervised training method.
10. The apparatus of claim 9 , wherein the classifier is to classify the change as representing one of a state of vacancy, a major motion, and a minor motion, wherein the space lighting system is deactivated during the state of vacancy.
11. The apparatus of claim 10 , wherein the classifier determines a confidence value associated with the classification.
12. The apparatus of claim 11 , further comprising a visible light sensor to measure an intensity of total visible light, wherein the intensity of total visible light is to be used by the classifier to determine the confidence value.
13. The apparatus of claim 12 , wherein the confidence value is used to control the space lighting system.