计算机工程与应用
計算機工程與應用
계산궤공정여응용
COMPUTER ENGINEERING AND APPLICATIONS
2013年
20期
150-152,221
,共4页
自动对焦%评价函数%对焦策略%感兴趣区域
自動對焦%評價函數%對焦策略%感興趣區域
자동대초%평개함수%대초책략%감흥취구역
auto focusing%evaluation function%focusing strategy%region of interest
为满足超大视场机器视觉自动对焦系统快速检测的要求，研究了一种基于对焦深度的自适应快速对焦算法。该算法采用粗对焦与精确对焦相结合的二级对焦策略，在不同的检测区域实现自适应选择最佳评价函数，实现自适应快速对焦。实验结果表明该算法在保持调焦精度的前提下，能够在大视场内快速聚焦，最终实现平均对焦时间小于0.5 s，一次性对焦成功率高达95%，满足自动对焦系统对高速、高精度自动对焦的要求。
為滿足超大視場機器視覺自動對焦繫統快速檢測的要求，研究瞭一種基于對焦深度的自適應快速對焦算法。該算法採用粗對焦與精確對焦相結閤的二級對焦策略，在不同的檢測區域實現自適應選擇最佳評價函數，實現自適應快速對焦。實驗結果錶明該算法在保持調焦精度的前提下，能夠在大視場內快速聚焦，最終實現平均對焦時間小于0.5 s，一次性對焦成功率高達95%，滿足自動對焦繫統對高速、高精度自動對焦的要求。
위만족초대시장궤기시각자동대초계통쾌속검측적요구，연구료일충기우대초심도적자괄응쾌속대초산법。해산법채용조대초여정학대초상결합적이급대초책략，재불동적검측구역실현자괄응선택최가평개함수，실현자괄응쾌속대초。실험결과표명해산법재보지조초정도적전제하，능구재대시장내쾌속취초，최종실현평균대초시간소우0.5 s，일차성대초성공솔고체95%，만족자동대초계통대고속、고정도자동대초적요구。
In order to meet the requirements of fast autofocus in focusing systems for rapid detection, a new adaptive focusing algorithm based on DFF is put forward. The algorithm uses coarse focus and precise focus with the two adaptive focusing strate-gies in different regions to achieve adaptive search to select the best evaluation function aiming to achieve autofocus. Experimental results show that the algorithm to maintain the premise of focusing accuracy can quickly focus on a large scale, and choose areas of interest to achieve accurate focus locally, and ultimately the average focusing time is less than 0.5 s, the success rate of one-time focus is 95%, to meet the focusing system for high-speed, high-precision AF requirements.
