本文針對現有技術中存在的問題提出一種對復雜背景下的多個移動物體進行目標快速識別與跟蹤的復合算法。該算法中采用對連續(xù)圖像進行差影計算來確定移動目標區(qū)域，從而能去除復雜背景干擾，可以明顯提高目標識別的速度和準確率。該系統已被實驗證明其有效性和實用性，可廣泛應用于監(jiān)測與識別系統，也可應用于無人監(jiān)控，無人自主操作等各種領域。本文中所提到的算法現為國家某重點項目中的關鍵技術。
關鍵字：差影算法圖像分割閾值 連通線 灰度值；
中圖文分類號TP274
Abstract: In this paper, we mention a multiple arithmetic in allusion to the problem of existing technology which can recognize and track the target among many motive objects in complex background. This arithmetic can determinate the area of motive target through the difference calculating of consecutive image, which can get rid of the interference of complex background, then improve the speed and exactitude of recognizing target obviously. This system has been proved efficiency and utility by test, which can broadly apply not only in the system of inspection and recognition but also the realm of self-control and self-operation etc.
Key words: different shadow arithmetic/image segmentation/threshold value/consecutive line/gray degree;