查詢結果分析
來源資料
頁籤選單縮合
題名 | 手眼協調攝影機在物體形狀辨識與抓取之研究=Recognition and Grasp of an Object Using a Hand-Eye Coordinated Camera |
---|---|
作者 | 呂明山; | 書刊名 | 明志技術學院學報 |
卷期 | 31 1999.07[民88.07] |
頁次 | 頁35-45 |
分類號 | 448.94 |
關鍵詞 | 手眼; 辨識; 指紋特徵; 傅立葉轉換; 位置基礎; 校正; Hand-eye; Recognition; Signature feature; Fourier transforms; Position-based; Calibration; |
語文 | 中文(Chinese) |
中文摘要 | 本文主要是應用固定式的手眼協調攝影機作為工作物辨識及機械手臂定位抓取工 作物的研究。結合影像的指紋特徵及傅立葉轉換的振幅分佈圖作為工作物外形的特徵,此方 法的特點為對於不同擺向與大小而有相同形狀的物體可獲得相同特徵。接著引用最臨近分類 法來辨識物體。在機械手臂定位方面,引用位置基礎的技術,其中包括影像轉換、手眼的校 正及座標轉換。此攝影機械手臂整合系統可應用於生產線上任意擺置物體的辨識分類與抓取 的工作。 |
英文摘要 | In this paper, the recognition and grasp of an object with a hand-eye coordinated camera is developed. Combination of signature feature and the spectrum of Fourier Transforms is used to represent the object's characteristic. It provides same feature for the objects with same shape but different orientation and size. The recognition of objects is achieved by using the method of Nearest Neighbor Classification. For the positioning of a robot, a position-based technique is introduced. It involves image transformation, hand-eye calibration and coordinate transformation. This hand-eye coordinated system developed can be implementated into the recognition and grasp of an object randomly posed in the production line. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。