System of optical noncontact microtopography

Optical triangulation-based microtopographic inspection of surfaces

The non-invasive inspection of surfaces is a major issue in a wide variety of industries and research laboratories. The vast and increasing range of surface types, tolerance requirements and measurement constraints demanded during the last decades represents a major research effort in the development of new methods, systems and metrological strategies. The discreet dimensional evaluation the rugometric characterization and the profilometric inspection seem to be insufficient in many instances. The full microtopographic inspection has became a common requirement. Among the different systems developed, optical methods have the most important role and among those triangulation-based ones have gained a major status thanks to their flexibility, reliability and robustness. In this communication we will provide a brief historical review on the development of optical triangulation application to the dimensional inspection of objects and surfaces and on the work done at the Microtopography Laboratory of the Physics Department of the University of Minho, Portugal, in the development of methods and systems of optical triangulation-based microtopographic inspection of surfaces.

Automation of a series of cutaneous topography measurements from silicon rubber replicas

BACKGROUND/AIMS

Skin relief is a matter of interest for dermatologists and surgeons. One of the methods available for surface topography measurement is based on 3D profilometry using skin surface replicas. and most studies use statistical results obtained from a large number of skin replica samples. The advent of optical profilometers (without contact) made it possible to remove the solid positive replica and to reduce the duration of the profilometric data acquisition. Nevertheless this saving of time, to be really interesting, needs to automate the data acquisition on a series of negative replicas.

METHODS/RESULTS

By adding a video camera to the optical profilometer and then by processing the resulting images, we have conceived a system able to carry out topographic measurements on a series of replicas loosely organized on a sample holder, without any human intervention. The silicon replicas in use have a very light colour: nearly white, sometimes slightly blue or green. The laser spot of the profilometer is so luminous that its red colour looks white through the camera. When choosing a replica holder with a matt dark colour and marking the left upper corner of the study area on the replica in black ink, the colours to be differentiated on the image are then close to the black one and the white one. We accordingly change the colour camera image into a black and white image (with 256 grey levels) and then carry out thresholdings to separate the different objects or information included in this image. With the use of a perfectly circular replica, of an accurately known size, laid on the sample holder at the center of the area filmed by the camera, we adjust the threshold level, which allows separation of the replica from its holder. We then move this calibrated replica in order to find the relationship between the size in pixels and the real size on the sample holder, in various positions of the video image. The software has four main built-in stages: Moving the sample holder beneath the sensor until a part of a replica is detected in the field of view of the camera; Moving the sample holder until this replica lies just in the middle of the image given by the camera; Recognition of the mark of the upper left corner of the surface area to be measured out inside this replica; and Moving the sample holder until the laser spot of the profilometer coincides with the origin of the surface area to be measured out, then carrying out this measurement. From the upper left corner of the sample holder, a scanning, line-by-line or column-by-column (according to the selected priority direction), is carried out until the successive replicas are found, and is stopped as soon as the number of replicas entered by the operator is reached.

CONCLUSION

The simplicity of the algorithms used makes it possible to distinguish the next measurement area from the preceding one in a few seconds.