基于视觉显著性的农作物图像评价方法研究

Study on Crop Image Evaluation Methods Based on Visual Saliency

近年来,随着电子信息技术、计算机可视化技术和互联网技术的迅速发展,传统的农业管理方式正逐渐被新型的农业信息化管理方法所替代。当前,基于数字图像处理技术的智慧农业已经成为新型农业信息化中的关键研究领域。具体来说,利用数字平台获取的农作物图像可以为专家提供非常丰富的信息,比如,农作物的长势,农作物病虫害情况等。但是,值得注意的是,获取这些有价值信息的前提是数字平台拍摄的农作物图像具有足够的清晰度,即没有出现严重失真。基于此,以农作物图像的质量为研究对象,提出了一种基于视觉显著性的半参考质量评价方法,在梯度域提取了方向直方图特征来刻画图像质量的变化。实验结果表明,该方法能够很好地识别农作物的图像质量,保证了后续高层次信息提取的有效性。

Along with the rapid development of electronic information technology,computer visualization technology and Internet technology,the traditional mode of agricultural management is gradually replaced by a new type of agricultural information.At present,intelligent agriculture based on digital image processing technology has become a key research field in the new agricultural informatization.In particular,crop images obtained from digital platforms can provide experts with a wealth of information,such as the growth trend of crops,crop diseases and insect pests.However,it is important to note that these valuable information is the premise of digital platform of crop the image clear enough,namely no serious distortion.Therefore,taking the quality of crop images as the research object,we propose a semi-reference quality evaluation method based on visual saliency.The direction histogram features from the gradient domain are extracted to describe the change of image quality.The experiment shows that the proposed method can well identify the quality of crop images and ensure the follow-up.