【摘要】 目前国内对于鸡胚生物反应器的成活性检测主要依靠人工照检，存在劳动强度大、效率低等缺点，提出一种仿生胚蛋成活性图像无损检测方法。针对以往研究中胚蛋血管与胚体区背景灰度差距较小，影响胚蛋血脉提取问题，提出一种基于双线性插值的限制对比度自适应直方图均衡算法对去噪后的胚蛋图像进行图像增强，基于匹配滤波对鸡胚血管进行分割，再通过连通域去噪完成鸡胚主血脉的二值形态的构建，根据鸡胚的颜色、气室、血管特征建立决策树分类器，来完成不合格鸡胚的剔除。结果表明，每枚鸡胚识别时间为0.634 s，不合格鸡胚判别的准确率为100%，满足疫苗生产的需求。更多还原

【Abstract】 At present,the viability detection of chicken embryo Bioreactor in China mainly relies on manual examination,which has the disadvantages of high labor intensity and low efficiency. In order to solve the problem of extracting the blood vein of embryo egg,the gray level difference between the blood vessel of embryo egg and the background gray level of embryo body region is small,in this paper,an adaptive histogram equalization algorithm based on Bilinear interpolation contrast restriction is proposed to enhance the de-noised embryo image,then,the binary morphology of the main blood vessel of chicken embryo was constructed by denoising the connected domain,and the decision tree classifier was established according to the color,gas chamber and vascular characteristics of chicken embryo. The results showed that the recognition time of each chicken embryo was 0.634 s,and the accuracy rate of unqualified chicken embryo was 100%,which could meet the needs of vaccine production.更多还原