Abstract
If you want to select an infrared camera, it is extremely important in order to better the attributes of these cameras, which most influence on the quality infrared images that are produced to understand. This document covers the three areas that influence picture quality: thermal pixel resolution, sensitivity and thermal noise sample sets. Each region has a significant impact on the quality of the thermal image.
If you bought a digital camerathe past may have influenced your purchase of your belief that the number of pixels in the main specification for the assessment of image quality for all decisions made by cameras. To read all of Consumer Reports ™ and detailed assessment of digital cameras, you'll enjoy this performance of the camera includes a thorough analysis of much more than the number of pixels. For a thermal imaging camera is essentially an image intensifier (radiant heatthe visible image), you should understand what are the key attributes to determine the quality of the thermal image and how they all contribute to the picture quality you may face, in your application.
Resolution in pixels The first consideration is the number of pixels. Today, there are three standard resolutions (some camera manufacturers' slightly) differ:
Low resolution – 160×120 (19600 points)
Medium Resolution – 320×240 (76800 points)
TopResolution – 640×480 (307,200 pixels)
As the resolution you need (verses) is mainly determined by the software and the value that you give a good picture. In evaluating a digital camera of 5 megapixels Verses 10 Most users will never benefit by buying a camera with 10 megapixels, for they will never print images on paper large enough that resolution by providing a superior print quality. Given that you will always display and print the fullResolution of an infrared camera for the highest resolution is relatively low compared to other digital cameras today. Even at a resolution of 640×480 pixel images in high definition thermal only a fraction of the computer screens to choose from today and the quality of the resulting thermal image printing will always be fully realized. Therefore, the resolution in the evaluation of a thermal imaging camera, the number of pixels is increased is relevant and the most important factor inImprovement of image quality.
Another advantage of the high resolution is the ability to enlarge a scene, and maintain a good image quality. The majority of the thermal cameras have an optical standard with a horizontal picture angle of 25 °. Apart from the leadership of a 640×480 pixel camera with a value of 2X digital zoom can be the performance of a camera with a resolution of 320×240 match option (and often expensive) 12 ° (2X) lens. If you have to assumefor an image of objects at a distance of 20 yards further than you were the increased cost of a lens 2X to a thermal camera 320×240 when comparing the total cost of between 320×240 and 640×480 systems.
The second major problem that affects the image quality is its sensitivity to heat. Although it uses a series of tests to this specification should be quantified, defined essentially as the thermal sensitivity of the camera, as you increase the image contrast. Thermal sensitivity varieswith the temperature object, such as increasing the temperature of the object increases, the slope increases the output signal of the detector with the temperature. This means that improved (increase) noise signal (solid) ratio, as you see the hot objects. However, it is usually not an advantage for applications where heat sensitivity can be better used are low temperature (room temperature) applications where the thermal contrast (delta temperature) in an image is very low. LowThe applications include the construction of the thermal contrast diagnosis when the imaging camera on the interior walls with variations in temperature or very small differences in the emission and topics such as moisture or insulation quality can be displayed in increasing the contrast at the point where the thermal camera sensitivity of the scope of the relevant parameters in temperature.
If you study the specifications, camera, enter the data to see thermal sensitivity of 0.25 ° C(250mK) and 0.05 ° C (50mK). While you might consider to fourth degree, adequate heat sensitivity when viewing a scene is limited, however, find the picture quality of a negative impact on image quality than the noise begins to dominate the picture.
Thermal imaging cameras are mostly pictures of the palette consists of 256 discrete colors or shades of gray. Make your goal is in front of a difference in temperature between 0 ° C and 256 ° C depending on grayscale or color represent 1 degreeTemperature difference. Now we turn the image of the same color in a scene with temperatures between 25 ° C and 35 ° C or 10 degrees. Each color is now 0.03 ° C (10 ° C ÷ 256), a lower value than the sensitive uncooled cameras. The result is a screen noise. There are many applications where it is very important to see the scope as close as possible to define the smallest possible temperature fluctuations. If you have a camera with a sensitivity of 0.25 ° C and wasget the same noise you a temperature of 65 ° C (150 ° F), which would probably lead to a very low image contrast must be set. They should recognize the difference between a camera with 50mK sensitivity for a camera with 100mK sensitivity was 100% and not more than 0.05 ° C better.
Thermal sensitivity
NETD is the temperature difference is just around the scene, internal noise of the detector (sensor NETD), or a totalelectronic noise measurement system (system NETD). As a buyer, you need a camera to evaluate the system NETD. The test consists of the assembly reference control temperature black body and a kind of environment (passive), which creates an easy target slot to see the camera. The black set temperature until it is almost equivalent to the ambient target. An oscilloscope measures the analog video output of a horizontal line and the point where theDelta temperature between the reference and objectives Netd room has a measurable signal is measured by the temperature difference between the baseline and benchmarks space.
MRTD – minimum resolvable temperature difference
This is a test system. An observer is invited to the difference in the temperature at which at least one target of 4 bars by the video output than the indicated temperature setpoints can be solved ratingReference Room and objectives agreed. This minimal difference will change with the spatial frequency of the target bars. A curve is obtained MRTD against spatial frequency, which characterizes the performance of the imaging system. Infrared systems have a modern low MRTD spatial frequency of several tens of milli-Kelvin.
The advantages of large format cameras is important for us to combine the need for a high sensitivity, while high spatialFrequencies.
To simplify the explanation of the basis of thermal sensitivity we focus on a single pixel sensor in an infrared uncooled infrared camera. Each pixel of a sensor focal plane array uncooled image is produced essentially a resistor with MEMS (micro-electro-mechanical systems ).
The basic structure of a pixel uncooled thermal imaging camera is disposed a bridge structure in the microscopic resistance of a material and a thin layer of absorbent were.Legs suspension bridge deck using an integrated circuit and provide electrical connection between the resistor bridge circuit and the reading of silicon. Reading IC, that the bias voltage, the resistance of thin layers, and multiplexing the signals of all pixels in the electronic imaging camera controls.
The form of infrared radiation by the temperature of each pixel changes as the energy of photons absorbed (8-14 micron wavelength) is converted into heat, which in turn changes the resistance of theThin-film resistor pixels. Reading IC sends a voltage to each bolometer "micro" component and a signal proportional to the heat absorbed by each detector is the basis of a video image in real time.
The electric circuit of an infrared sensor is very easy to activate voltage for each pixel, and a change in resistance, the resistance of thin films based on the temperature and a pixel is scanned into a numeric value. All analog signals have a certain level of noise with the signalgenerated by the sensor. The signal to noise ratio significantly affect the quality of the camera, because the noise level is usually a fixed amount and the gain of the detector increases the system start to the signal-to-screen and start to see the "snow" in the picture.
The level of this noise is usually specified as a noise equivalent temperature difference.
As with any electrical circuit, there are a multitude of possibilities for electrical interference from enteringSystems, but the quality (signal to noise ratio) of the input signal directly to the infrared pixel has the greatest influence on the thermal sensitivity, as almost all developers to create the camera on the same electronic components, a device that photo. Therefore, the thermal sensitivity is largely on the quality of the IR imager array is based.
Other issues such as the number f of the lens and thermal sensitivity. Your camera's lens should ƒ1.0 (theFocal length is equal to the diameter of the target) is considered a "fast" lens. In comparison, the number of your digital camera between ƒ3 f ƒ5, and probably during the cameras used in mobile phones and other low-cost systems can be as high as ƒ20! As conditions for an application to longer focal lengths, it is useful to slow down on "approach to reduce the size, weight and cost of lenses and compromise in some thermal sensitivity. For example, a F1.4 optic leads2X reduction in thermal sensitivity and a 4x optical F2.0 reduced thermal sensitivity. Therefore, a system with 50mK sensitivity with a normal lens, a good sensitivity (100mK) for a tele-ƒ1.4 be maintained in order to attached the camera to another device, the thermal sensitivity began to 100mK and 200MK is on display through a " slower (ƒ number greater than 1).
As you can see on the various issues in this paper, the nature of the issues raisedthermal sensitivity is very complex, but in the real world of the human eye is very easy to distinguish small differences in image quality can be found from (high sensitivity), if you see him.
Non-uniformity correction
Since the number of pixels increases sensitivity and improves the image quality always depends on a process called non-uniformity calibration or NUC. As we have described a series of microbolometer imaging is essentially aTable of tiny resistors, and because of the micro-scale of these devices, there are variants, as each pixel reacts to infrared energy from an object. Be standardized in the manufacture of the infrared camera's sensor has to be, which means that differences in the response and DC output for each sensor must be set to zero. Thermal cameras usually have an internal flag or iris, which is regularly put in front of the sensor as a constant temperature of zero referenceDifferences between the pixels. This is a fine tuning of the process of NUC and the plant is sometimes a button "up" mentioned.
As processing source inside the lens, and further improvements in image quality are possible if you edit calibration through the lens with a lens cover or expose the unit to a large uniform surface. When will not improve flight performance-uniformities created by the lens start to see and for the final imagethe quality of the simple one-step lens calibration ensures the best image quality of the camera is able to generate.
Advantages of high image quality:
Much more flexibility to review the objectives are too variable distances
Ability goals prospect of low thermal contrast
More intuitive diagnoses problems caused by heat
Visible Infrared Enhanced image quality was melting due to the better resolution of the camera in accordance with infrared and visible ..
Flexibilityaccept a lower cost and less weight, lens options
Diagnoses intuitive temperature anomalies