Abstract
— In this paper, we proposes a machine vision system to approximate depth of drilled dental plaster model in training process of young dentist and undergraduate dentist. The system is based on basic laser scanner and machine vision technique. The proposed system consists of two main parts: an image acquisition process using basic line laser scanner and depth measurement system based on a right and isosceles triangle. 450 laser points on each dental plaster model sample are extracted for testing. From our experimental result, our system is more effectiveness to inspect the practiced dental plaster model than manual inspection by a senior dentist. Our efficiency proposed system is 0.1 mm error in depth on measurement range 0.2 to 1.2 mm depth.
Keywords-laser beam; depth estimation; skeleton; 3D model;dental drill machine; machine vision;
I.
INTRODUCTION
A dental drill is an important and basic tool for the dentist in dental treatment; remove infected and polish from teeth [1]. In skill training procedure, dental students are trained by dental plaster models with the dental drill. An inspector, senior dental, rates their skill by measuring drilled plaster models with a periodontal probe in 1 mm. scale. This process takes a long time. It is subjective measurement relying on inspector’s experience, physical condition, and others. An automatic surface inspection by machine vision system is role in industrial very fast. Many advantages of machine vision system are non-destructive testing, contactless, high reliability, 100% inspection, real-time working. Human vision knowledge is easily applied to the system. Normally, surface inspection system integrates two main parts: light source system (light bulb and laser beam) and machine vision algorithms [2-10]. For the light bulb, a single light bulb for road surface inspection by using reflected surface characteristics [2], multi-color light bulbs for rail surface inspection by using color surface characteristics [3]. Laser beam is applied for high resolution and high quality applications. One or more laser colors and one or more digital cameras are applied to increase performance [4, 5], a mobile car laser scanner for road surface inspection system [6,7], surface of metal steel inspection system which detects shiny and smooth surface of iron sheet [8,9], sphere defect detection by using 3D model [10], and others. These applications help to improve speed, precision, and accuracy of inspection system.
However, some system is complicated and expensive. In this paper, we present a simple dental plaster model inspection system. In Section II presents inspection system prototype and image acquisition, followed by our implemented system in Section III, and finally experimental result in Section IV.
dental laser tips
II.INSPECTION SYSTEM AND IMAGE ACQUISITIONA.Inspection System Prototype Distinguished laser properties are very high luminance and narrow band, which is not been disturbed by other light sources and reflection patterns: diffuse, nearly ideal specular and mixed diffuse, and specular reflection [11, 12]. In our system, we design a simple system which is based on basic laser scanner ideal using a line laser beam as a light source shown in Fig. 1.
shows efficiency of our system. 450 testing points from the dilled dental plaster model are measured by dual gauge for 'real depth measurement' and our introduced system for “depth approximation”. Error rates of the real depth measurements and depth approximations are compared and shown in Table I. In term of effectiveness measurement is shown in Table II. Our introduced system works faster than an expert inspection with 3 times.
In this paper, we introduced an automatic system to inspection dental plaster model examples for young dental practice and undergraduate dentist by basic laser scanner and machine vision techniques. The system reduces amount of verifying and scoring the practice dentals and also provide high accuracy. Future work, variation of dental practices such as smooth drilled, curve drilled, sine wave drilled, and others will be integrated to create a method to measuring and scoring model for young dental practice scoring.
This research is support by faculty of engineering, Prince of Songkla University (PSU). We would like to acknowledge all suggestions from all researchers at intelligent system laboratory (iSys), department of computer engineering, PSU.