Skip to main content
. 2016 Jun 30;16(7):1010. doi: 10.3390/s16071010

Table 2.

Description of 12 datasets.

Dataset Condition Detail Description
I (see in Figure 11a) 1.2 °C, morning, humidity 73.0%, wind 1.6 m/s
  • -

    The intensity of background is influenced by the window of building

  • -

    Human shadow is reflected on the window, which is detected as another object in thermal image

II (see in Figure 11b) −1.0 °C, evening, humidity 73.0%, wind 1.5 m/s
  • -

    The intensity of background is influenced by the window of building

  • -

    Object is not seen in visible light image

  • -

    Human shadow is reflected on window, which is detected as another object in thermal image

III (see in Figure 11c) 1.0 °C, afternoon, cloudy, humidity 50.6%, wind 1.7 m/s
  • -

    The intensity of background is influenced by leaves and trees

IV (see in Figure 11d) −2.0 °C, dark night, humidity 50.6%, wind 1.8 m/s
  • -

    The intensity of background is influenced by leaves and trees

  • -

    Object is not seen in visible light image

V (see in Figure 11e) 14.0 °C, afternoon, sunny, humidity 43.4%, wind 3.1 m/s
  • -

    Difference between background and human diminishes because of the high temperature of background

VI (see in Figure 11f) 5.0 °C, dark night, humidity 43.4%, wind 3.1 m/s
  • -

    The air heating system of building increases the temperature of part of the building in background

  • -

    Object is not seen in visible light image

VII (see in Figure 11g) −6.0 °C, afternoon, cloudy, humidity 39.6%, wind 1.9 m/s
  • -

    Halo effect is shown near the human area in thermal image, which makes it difficult to detect the correct human area

VIII (see in Figure 11h) −10.0 °C, dark night, humidity 39.6%, wind 1.7 m/s
  • -

    Halo effect is shown near the human area in thermal image, which makes it difficult to detect the correct human area

  • -

    Object is not seen in visible light image

IX (see in Figure 11i) 21.9 °C, afternoon, cloudy, humidity 62.6%, wind 1.3 m/s
  • -

    Halo effect is shown near the human area in thermal image, which makes it difficult to detect the correct human area

  • -

    Difference between background and human diminishes due to the high temperature of background

X (see in Figure 11j) −10.9 °C, dark night, humidity 48.3%, wind 2.0 m/s
  • -

    The dataset was collected at night during winter. Therefore, the background in thermal image is too dark because of low temperature

  • -

    Object is not seen in visible light image

XI (see in Figure 11k) 27.0 °C, afternoon, sunny, humidity 60.0%, wind 1.0 m/s
  • -

    Human is darker than road because the temperature of the road is much higher than that of a human in summer

  • -

    Leg is not clear when kicking behavior happens because the woman in the image wore a long skirt

XII (see in Figure 11l) 20.2 °C, dark night, humidity 58.6%, wind 1.2 m/s
  • -

    Human is darker than road because the temperature of the road is much higher than that of a human in summer

  • -

    Object is not seen in visible light image