Object recognition by far infrared camera
Example implementations described herein are directed to integration of far infrared cameras in a vehicle system to detect objects based on relative temperature of objects. Such implementations can improve accuracy when paired, for example, with classification systems that classify objects based on the shape of the object, as both the shape and relative temperature can be used to ensure that the classification is accurate. Further, example implementations can synchronize far infrared cameras with other sensor systems to determine distance, energy, and absolute temperature of an object, which can also be used to enhance classification. Such classifications can then be provided to an advanced driver assistance systems (ADAS), which can control the vehicle system in accordance with the object classification.
1. A vehicle system, comprising:
one or more far infrared (FIR) cameras; and
a processor, configured to:
classify a first object detected from one or more images of the one or more FIR cameras;
for a second object detected from the one or more images of the one or more FIR cameras having a relative temperature range coinciding with the first object, recognize the second object as having the classification of the first object;
determine an energy range of the second object based on the classification of the first object; and
determine a distance range to the second object based on the energy range of the second object and the relative temperature range.
2. The vehicle system of claim 1 , wherein the processor is configured to classify the first object detected from the one or more images of the one or more FIR cameras by:
classify the first object detected from the one or more images of the one or more FIR cameras based on a shape of the first object;
determine a relative temperature range of the first object; and
associate the relative temperature range of the first object to the classification of the first object.
3. The vehicle system of claim 1 , further comprising a sensor system configured to provide image information and distance information, wherein the processor is configured to synchronize the one or more images of the FIR cameras with the image information and distance information from the sensor system, and classify the first object based on the image information and distance information.
4. A vehicle system, comprising:
a thermometer configured to measure a temperature of a point in proximity to the vehicle system;
one or more far infrared (FIR) cameras; and
a processor, configured to:
classify a first object detected from one or more images of the one or more FIR cameras;
for a second object detected from the one or more images of the one or more FIR cameras having a relative temperature range coinciding with the first object, recognize the second object as having the classification of the first object;
determine an absolute temperature range of the first object based on the temperature of the point in proximity of the vehicle system and the relative temperature range of the first object;
for the second object detected from the one or more images of the one or more FIR cameras:
process overlapping ones of the one or more images from multiple ones of the one or more FIR cameras;
convert the relative temperature range of the second object into an absolute temperature range of the second object based on the temperature of the point in proximity to the vehicle system and the relative temperature range of the second object; and
for the second object having the absolute temperature range of the first object, recognize the second object as having the classification of the first object.
5. The vehicle system of claim 1 , wherein the processor is further configured to:
determine energy for the first object based on temperature and a distance to the first object;
classify the first object detected from one or more images of the one or more FIR cameras based on the energy; and
for the second object detected from the one or more images of the one or more FIR cameras having a relative temperature range coinciding with the first object and within an energy range of the first object, recognize the second object as having the classification of the first object.
6. The vehicle system of claim 1 , further comprising an advanced driver assistance system (ADAS) configured to control the vehicle system according to the classification of the first object and the second object.
7. The vehicle system of claim 1 , the processor further configured to:
for the second object detected from the one or more images of the one or more FIR cameras not having the relative temperature range coinciding with the first object, recognize the second object based on the relative temperature range in comparison to relative temperature ranges of previous detected objects.
8. A non-transitory computer readable medium, storing instructions for executing a process for a vehicle system comprising one or more far infrared (FIR) cameras, the instructions comprising:
classifying a first object detected from one or more images of the one or more FIR cameras;
for a second object detected from the one or more images of the one or more FIR cameras having a relative temperature range coinciding with the first object, recognizing the second object as having the classification of the first object;
determining an energy range of the second object based on the classification of the first object; and
determining a distance range to the second object based on the energy range of the second object and the relative temperature range.
9. The non-transitory computer readable medium of claim 8 , wherein the classifying the first object detected from the one or more images of the one or more FIR cameras comprises:
classifying the first object detected from the one or more images of the one or more FIR cameras based on a shape of the first object;
determining a relative temperature range of the first object; and
associating the relative temperature range of the first object to the classification of the first object.
10. The non-transitory computer readable medium of claim 8 , the vehicle system further comprising a sensor system configured to provide image information and distance information, wherein the instructions further comprises synchronizing the one or more images of the FIR cameras with the image information and distance information from the sensor system, and classifying the first object based on the image information and distance information.
11. A non-transitory computer readable medium storing instructions for executing a process for a vehicle system, the vehicle system comprising a thermometer configured to measure a temperature of a point in proximity to the vehicle system, and one or more far infrared (FIR) cameras, the instructions comprising:
classifying a first object detected from one or more images of the one or more FIR cameras;
for a second object detected from the one or more images of the one or more FIR cameras having a relative temperature range coinciding with the first object, recognizing the second object as having the classification of the first obi ect;
determining an absolute temperature range of the first object based on the temperature of the point in proximity of the vehicle system and the relative temperature range of the first object;
for the second object detected from the one or more images of the one or more FIR cameras:
processing overlapping ones of the one or more images from multiple ones of the one or more FIR cameras;
converting the relative temperature range of the second object into an absolute temperature range of the second object based on the temperature of the point in proximity to the vehicle system and the relative temperature range of the second object; and
for the second object having the absolute temperature range of the first object, recognizing the second object as having the classification of the first object.
12. The non-transitory computer readable medium of claim 8 , the instructions further comprising:
determining energy for the first object based on temperature and a distance to the first object;
classifying the first object detected from one or more images of the one or more FIR cameras based on the energy; and
for the second object detected from the one or more images of the one or more FIR cameras having a relative temperature range coinciding with the first object and within an energy range of the first object, recognizing the second object as having the classification of the first object.
13. The non-transitory computer readable medium of claim 8 , the instructions further comprising:
for the second object detected from the one or more images of the one or more FIR cameras not having the relative temperature range coinciding with the first object, recognize the second object based on the relative temperature range in comparison to relative temperature ranges of previous detected objects.