Image enhancement for crop trait information acquisition system

Highlights

• Overview of the previously proposed crop trait information acquisition system.

• An adaptive image enhancement method based on guided filtering is proposed.

• An enhancement weight based on image illumination intensity was developed.

• The proposed method yielded visually better results than several traditional methods.

• Objective evaluation measures confirmed the improved performance of our model.

Abstract

Collecting images using portable devices is an effective and convenient method for acquiring crop trait information. Because of uncertain environmental conditions in the field, enhancement is necessary to improve the visual quality of images. With this aim, here we propose an adaptive image enhancement method based on guided filtering. Our method automatically calculates the enhancement weights of the detail in an image according to the distribution characteristics of the illumination intensity of a crop image, so as to adaptively adjust the contrast of the image. To verify the effectiveness of the proposed algorithm, we performed enhancement experiments on 50 images of four kinds of cucumber leaf tissues, namely, leaves infected with target spot, powdery mildew, and downy mildew, and healthy leaves. The results showed that our proposed method substantially improved the visual quality of the images. Moreover, the mean ratios of the contrast to color difference obtained using the proposed method were higher than the mean ratios obtained using five conventional enhancement methods. We consider the proposed method for image enhancement will be a valuable addition to the crop trait information acquisition system (http://ebreed.com.cn/).