CAJ | 학술논문

针对红外与可见光图像特征融合提出一种改进免疫遗传算法的融合方法。针对传统免疫遗传算法的不足，存在的早熟收敛，搜索过程缓慢，以至于种群进化停滞不前等缺点，将自适应和“淘汰保留”方法应用到免疫遗传算法中。通过特征编码，种群初始化，应用改进的免疫遗传算法进行特征融合。通过实验仿真结果证明改进的免疫遗传算法有较快的收敛速度和较好的搜索能力，使用基于改进的免疫遗传算法的图像融合特征具有较好的识别率。
침대홍외여가견광도상특정융합제출일충개진면역유전산법적융합방법。침대전통면역유전산법적불족，존재적조숙수렴，수색과정완만，이지우충군진화정체불전등결점，장자괄응화“도태보류”방법응용도면역유전산법중。통과특정편마，충군초시화，응용개진적면역유전산법진행특정융합。통과실험방진결과증명개진적면역유전산법유교쾌적수렴속도화교호적수색능력，사용기우개진적면역유전산법적도상융합특정구유교호적식별솔。
This paper proposes an improved immune genetic algorithm fusion method of infrared and visible light image charac?teristics. Because of the defects in the traditional immune genetic algorithm, the existence of the premature convergence, the search process is slow, so that the disadvantage of population evolution remain stagnant, adaptive and"out of retained"is applied to the immune genetic algorithm. Through the feature encoding, population initialization, the improved immune genetic algo?rithm is used for feature fusion. Through the experimental simulation results show that the immune genetic algorithm improved the convergence rate is faster and better search ability, the use of the improved immune genetic algorithm of image fusion based on feature has good recognition rate.