科技通报
科技通報
과기통보
BULLETIN OF SCIENCE AND TECHNOLOGY
2013年
9期
152-156
,共5页
蚁群算法%维吾尔文字图像%初步搜索%更新策略
蟻群算法%維吾爾文字圖像%初步搜索%更新策略
의군산법%유오이문자도상%초보수색%경신책략
ant colony algorithm%uygur text images%preliminary search%update strategy
针对传统蚁群算法在维吾尔文字图像分割时容易产生缺陷和干扰的缺点,提出了一种改进的蚁群算法。该算法首先利用区域蚁群通过聚类中心对图像进行初步搜索和分割;然后再引入边界蚁群通过不同的路径选择信息素和策略更新手法来对图像进行边界补偿搜索;最后利用两种蚁群分别采用不同的路径选择信息素和筻略更新,共同实现对维吾尔文字图像精确分割的目标。实验结果表明,该算法准确分割出维吾尔文字区域,提高了其识别精度。
針對傳統蟻群算法在維吾爾文字圖像分割時容易產生缺陷和榦擾的缺點,提齣瞭一種改進的蟻群算法。該算法首先利用區域蟻群通過聚類中心對圖像進行初步搜索和分割;然後再引入邊界蟻群通過不同的路徑選擇信息素和策略更新手法來對圖像進行邊界補償搜索;最後利用兩種蟻群分彆採用不同的路徑選擇信息素和筻略更新,共同實現對維吾爾文字圖像精確分割的目標。實驗結果錶明,該算法準確分割齣維吾爾文字區域,提高瞭其識彆精度。
침대전통의군산법재유오이문자도상분할시용역산생결함화간우적결점,제출료일충개진적의군산법。해산법수선이용구역의군통과취류중심대도상진행초보수색화분할;연후재인입변계의군통과불동적로경선택신식소화책략경신수법래대도상진행변계보상수색;최후이용량충의군분별채용불동적로경선택신식소화강략경신,공동실현대유오이문자도상정학분할적목표。실험결과표명,해산법준학분할출유오이문자구역,제고료기식별정도。
In view of the traditional ant colony algorithm in uygur text image segmentation easily when defects and interference, an improved ant colony algorithm is put forward. Using the algorithm of regional ant colony clustering center through the preliminary search and segmentation of image and Then introducing boundary ant colony through different path choice of pheromone and strategy update technique to image boundary compensation search;Finally by using two kinds of ant colony optimization, respectively by different path choice of pheromone and Gang slightly updated, achieve the goal of uygur text segmentation precise images together. The experimental results show that this algorithm can segment the uygur text area, improve the identification accuracy .