Patent Application
20120274778
Embodiments presented herein relate generally to infra-red imaging, and more specifically to transient thermographic inspection.
Thermographic inspection is the nondestructive testing of objects through imaging of thermal patterns on the object's surface. Thermographic inspection is often preferred to other nondestructive testing techniques such as ultrasonic inspection and radiographic inspection, for various advantages offered by thermographic inspection. Thermographic inspection is non-contact, non-intrusive, allows for detection of subsurface detects close to the surface, allows for inspection of large surfaces, and offers high speed inspection. One form of thermographic inspection is transient thermography. Transient thermography involves observing the temperature distribution on the surface of an object under test as it is subjected to a thermal transient such as a pulse of heat or a pulse of heat sink, and then allowed to return to ambient temperature. Any flaws present are detected as abnormalities in the surface temperature distribution during this thermal transient. Transient thermography is particularly well suited to the inspection of composite materials. The relatively low thermal conductivity of composite materials results in relatively long-lived thermal transients, therefore making the thermal transient easy to detect with a thermal camera.
Some known thermographic inspection techniques are operator dependent techniques, involving an operator to watch a thermal video of the object under test. The operator then observes the video for changes in contrast caused due to flaws within the object. Such techniques are skill intensive and require much manual effort. Automated thermographic inspection techniques also exist. Automated thermographic inspection uses a heat source such as a high intensity flash lamp to heat the surface of an object under test. An infra-red camera then takes a series of thermal images or thermograms of the object under test. The images are then post processed to identify features in the object under test.
Known automated thermographic inspection techniques use an infra-red camera driven at a constant frame rate. In other words, the infra-red camera is adapted to capture thermal images at fixed intervals of time. However, the thermal activity occurs at a non-linear rate. This causes too few thermal images being captured at the beginning of the thermal transient, when the thermal activity is high, and too many thermal images (with a large proportion being redundant images) being captured at the end of the thermal transient when the thermal activity is significantly slower. This causes very large data files that require a large buffer memory, a faster bus, and subsequently requires large disk space for storage and archival. These higher computing requirements increase the cost of the thermographic inspection system. Some techniques may partly address the storage problem by extracting thermal images from the large set of captured images. However, such techniques still require large buffers and fast busses to facilitate high speed image capture.
Therefore, there is a need for a thermographic inspection system that addresses these and other shortcomings associated with known solutions.
A method for thermographic imaging is described. The method captures a plurality of thermal images of a surface of an object at non-linear intervals over a period of time, each of the thermal images being associated with temporal data. The method further includes processing the plurality of thermal images and the temporal data, and identifying features within the object based on the processing.
A system for thermographic imaging is described. The system includes an infra-red imaging device that captures a plurality of thermal images of a surface of an object. The system further includes a variable time base generator that triggers the infra-red imaging device at non-linear intervals over a period of time. The system further includes a time module that associates temporal data with each of the plurality of thermal images. In some embodiments, the system may also include a processor for processing the plurality of thermal images taking into account the temporal data associated with each of the plurality of thermal images, and identifying features within the object. In some further embodiments, the system may also include a heater for heating the surface of the object.
Embodiments of a thermographic imaging system are presented herein. The system employs a non-linear time base for triggering an infra red (IR) camera, to enable imaging at a non-linear rate. In transient thermographic testing, for example, the object under test is heated by a predefined temperature. Heating is then stopped and the object under test allowed to return to ambient temperature. During the thermal transient, thermal activity of the object under test takes place at a non-linear rate dependant on various factors such as, but not limited to, emissivity, absorptivity, reflectivity, and temperature of the object under test. Specifically, the thermal activity occurs rapidly at the beginning of the thermal transient, and slows down as the thermal transient progresses. Thus, imaging at the non-linear rate may reduce the number of redundant images, thus reducing requirements of processing power, buffer memory, and image file sizes. The system stores temporal data associated with each of the captured images for further processing. In some embodiments, the system also includes a processing module for detecting and identifying features in an object under test.
The IR imaging device 102 may include an IR sensor such as, but not limited to, a focal plane array IR sensor. The IR sensor may be, for example, a vanadium oxide microbolometer array. It should be appreciated that other IR sensors may also be used. The IR imaging device 102 may further include one or more IR lenses to produce a sharp image on the IR sensor. The IR lens may be designed to pass only light in the infra red spectrum.
The frame buffer 104 temporarily stores the thermal images captured by the IR imaging device 102, until the thermal images can be transferred to a non-volatile storage device, or transmitted over a network to an image processing terminal. The frame buffer 104 may be a high speed semiconductor memory device.
The control module 106 controls the operation of the IR imaging device 102, and the frame buffer 104. The control module 106 includes the variable time base generator 108, and the time module 110. The variable time base generator 108 generates a non-linear time base signal for triggering the IR imaging device 102. The non-linear time base signal triggers the IR imaging device 102 at non-linear intervals of time. In other words, the variable time base generator 108 causes the frame rate of the IR imaging device 102 to change over time. In various implementations, the variable time base generator 108 triggers the IR imaging device 102 after short time intervals at the beginning of a thermal transient, where a noticeable difference in thermal activity occurs in a short time interval. As the thermal transient progresses, the variable time base generator 108 triggers the IR imaging device 102 after longer time intervals, since a difference in thermal activity may only be noticed at much longer time intervals. Thus, the use of a non-linear time base signal enables capturing an optimal number of thermal images and reducing the number of redundant thermal images.
The variable time base generator 108 may generate a continuously varying time base signal to trigger the IR imaging device 102.
The variable time base generator 108 may be implemented as an analog circuit using, for example, square law devices such as MOSFETs operating in the saturation region, capacitors for exponential time base, operational amplifiers and so forth. Alternatively, the variable time base generator 108 may be implemented as a digital circuit using, for example, a microprocessor, a programmable logic device such as a field programmable gate array, and so forth. In one example implementation, a microprocessor may be programmed with a time base computing function, to generate the variable time base signal.
In some low complexity thermal imaging systems, the variable time base generator 108 may be hardcoded or hardwired into the thermal imaging system 100. Such an implementation may be suitable for low cost thermal imaging systems and portable thermal imaging systems. Alternatively, the variable time base generator 108 may be a programmable or configurable time base generator. One example implementation of a programmable variable time base generator may accept user input of a time base function. In another example implementation, the programmable variable time base generator may have stored thereon in a memory, a plurality of time base functions. The programmable variable time base generator may accept user selection of one of the plurality of time base functions. Each of the plurality of time base functions may be associated with a specific type of object to be tested, or for specific materials, or for specific physical dimensions. Such a system may accept inputs to select the specific type of object to be tested, or the specific material, or the specific physical dimensions, or a combination thereof, and select a suitable time base function based on the input.
The control module 106 also includes a time module 110. The time module 110 generates temporal data for each of the captured thermal images. The temporal data may be an accurate time stamp precise to a millisecond, for example. The temporal data may then be used for processing the thermal images. In one implementation, the time module 110 stores the generated temporal data in the frame buffer 104, in association with the appropriate thermal image. The control module 106 may be configured to precisely synchronize the transfer of the thermal image captured by the IR imaging device 102 into the frame buffer 104, with the transfer of the temporal data from the time stamping module 110 to the frame buffer 104.
The image processor 114 then processes the thermal images stored in the frame buffer 104 to identify features within the object under test. The image processor 114 may use any suitable algorithm for processing the thermal images. However, the image processor 114 may now directly use the temporal data associated the thermal images to identify features, instead of the conventional technique of using frame numbers translated into an estimated time of thermal image capture.
At step 404, the IR imaging device 102 captures a plurality of thermal images of the surface over a period of time. The IR imaging device 102 being triggered by the variable time base generator 108, captures the plurality of thermal images of the surface at non-linear intervals of time. Further, each of the thermal images is associated with temporal data generated by the time module 110. The plurality of thermal images and the associated temporal data may be stored in the frame buffer 104.
At step 406, the image processor 114 processes the plurality of thermal images and the temporal data. In one example implementation, the image processor 114 determines the contrast of each pixel by subtracting a mean intensity of all pixels of that thermal image from the pixel intensity at that point of time. The contrast is then plotted versus time for each pixel, using the temporal data associated with the thermal images.
At step 408, the image processor 114 identifies features within the object, based on the processing. The image processor 114 may estimate a depth of the feature based on the temporal data associated with a peak of the contrast plot. Such identification is particularly useful where the feature's thickness does not force the heat to flow around the feature (also known as a “through heat” feature).
However, when the rate of diffusion of the heat is faster around the feature than through the feature (also known as a “lateral heat” feature), the contrast peak is not evident from the contrast plot. In such a case, the processor 114 computes, at step 406, the time derivative of the contrast plot to determine the time derivative peak of contrast. Subsequently, at step 408, the processor 114 may identify the depth of the “lateral heat” feature based on the temporal data associated with the time derivative peak of contrast.
Although specific implementations have been described in the embodiments presented herein, such implementations are exemplary, and should not be considered limiting in scope. Other variations of the implementations are also envisioned and the scope is defined by the appended claims.
