基於雷射光輔助檢測內螺紋之研究

Abstract

目前業界量測物件表面大廠包含Keyence及Memstec等之雷射式位移感測器，其感測透過固定雷射光源及CCD之兩點之夾角，並透過三角雷射計算出待測物表面高低之狀況，近而檢測出瑕疵物件之所在。由於上述之產品，其感測器本體大小，針對量測環狀物之內螺紋物件，易受限感測器本身尺寸而造成干涉狀況，導致無法量測物件。本論文採用雷射光源及CCD分離方式，不受限於待測物本身大小，以解決量測物件與感測器之干涉狀況。 本論文貢獻有三點，在2592×1944像素之圖像下，取用樣本298圖像，第一點，找尋出最佳HSV參數，牙差變化範圍在13.13%，約±76.97μm的誤差值。第二點，找出適用於穩定第一、二牙峰差值，為25個像素。第三點，在1920×1080像素之動態影像上，取用1260張動態圖像，找尋牙峰之邊緣及牙峰與牙底點為，穩定度在99.76%。

Keyence and Memstec are the main factory who major in object surface measurement currently including laser displacement sensor. It senses through the fixed laser light source and CCD’s two points of the angle, and also calculate the object’s surface height through the triangular laser to detect the defective object. Due to the sensor’s part body size of the above-mentioned products, the measurement which specify for the internal thread of a ring circuitloop-shaped object is easily to be interfered because of it’s size problem and resulting in the failure of object measurement. Therefore, we adapt the laser light source and CCD separation method, which are not limited by the size of the object itself to solve the measurement of objects and sensors of the interference problem. In this thesis, there are 3 contributions which access 298 image samples under the 2592×1944 pixel image. First, to find the best HSV of parameter, different of crest and root in the range of 13.13%, about ± 76.97μm error value. Second, find out the stabilizing of the first and two dentition difference for 25 pixels. Third, access 1260 dynamic images in the 1920×1080 pixel dynamic image, to find the edge of the crest and the stability of 99.76% between peak and the base.