Influence of Image Quality and Light Consistency on the Performance of Convolutional Neural Networks for Weed Mapping

Abstract

Recent computer vision techniques based on convolutional neural networks (CNNs) are considered state-of-the-art tools in weed mapping. However, their performance has been shown to be sensitive to image quality degradation. Variation in lighting conditions adds another level of complexity to weed mapping. We focus on determining the influence of image quality and light consistency on the performance of CNNs in weed mapping by simulating the image formation pipeline. Faster Region-based CNN (R-CNN) and Mask R-CNN were used as CNN examples for object detection and instance segmentation, respectively, while semantic segmentation was represented by Deeplab-v3. The degradations simulated in this study included resolution reduction, overexposure, Gaussian blur, motion blur, and noise. The results showed that the CNN performance was most impacted by resolution, regardless of plant size. When the training and testing images had the same quality, Faster R-CNN and Mask R-CNN were moderately tolerant to low levels of overexposure, Gaussian blur, motion blur, and noise. Deeplab-v3, on the other hand, tolerated overexposure, motion blur, and noise at all tested levels. In most cases, quality inconsistency between the training and testing images reduced CNN performance. However, CNN models trained on low-quality images were more tolerant against quality inconsistency than those trained by high-quality images. Light inconsistency also reduced CNN performance. Increasing the diversity of lighting conditions in the training images may alleviate the performance reduction but does not provide the same benefit from the number increase of images with the same lighting condition. These results provide insights into the impact of image quality and light consistency on CNN performance. The quality threshold established in this study can be used to guide the selection of camera parameters in future weed mapping applications.

Evaluating Cross-Applicability of Weed Detection Models Across Different Crops in Similar Production Environments

“…It should be noted that Cot2 differed from Cot1 in three aspects: (1) Cot2 had a relatively higher density of weeds and the median size of MG and Grass differed from that of Cot1, (2) some of the cotton plants in Cot2 had slightly different visual appearance due to herbicide drift, and (3) the illumination conditions for Cot2 was slightly darker than that of Cot1. Hu et al (2021) suggested that illumination conditions can affect weed detection accuracy. With respect to herbicide drift impact, Suarez et al (2017) found in cotton that drift can lead to a significant change in the spectral behavior of the crop.…”

“…Several image-based weed detection techniques have been proposed and implemented. Based on developments made so far, these techniques can be broadly categorized into two main groups: (1) traditional segmentation and machine learning-based techniques ( Wu et al, 2011 ; Ahmed et al, 2012 ; Rumpf et al, 2012 ; García-Santillán and Pajares, 2018 ; Sabzi et al, 2018 ; Sapkota et al, 2020 ) and (2) advanced computer vision using convolution neural networks (CNNs; Adhikari et al, 2019 ; Ma et al, 2019 ; Sharpe et al, 2020 ; Hu et al, 2021 ; Xie et al, 2021 ).…”

“… Yu et al (2019) employed DetectNet to detect dandelion ( Taraxacum officinale ), ground ivy ( Glechoma hederacea ), and spotted spurge ( Euphorbia maculata ) in perennial ryegrass. Hu et al (2021) tested Faster R-CNN, DeepLabv3, and Mask R-CNN for broadleaf and grass weed detection in cotton ( Gossypium hirsutum ) and soybean ( Glycine max ) using UAS-borne high-resolution images.…”

A Variable Radius Side Window Direct SLAM Method Based on Semantic Information

“…e camera overexposure addition operation in this study is similar to other pieces of literature [56]. e number of interval frames at which overexposure interference is added ranges from 30 to 10 and is decreased by a step size of 5. e results are shown in Table 9.…”

An Intelligent Solution for Automatic Garment Measurement Using Image Recognition Technologies

“…As one of the objectives of this study is to investigate the feasibility of automatic garment sizing with household photographs that can be uploaded to different platforms (e.g., second-hand clothing platforms) photos taken under different conditions and with different mannequins or hanging on a hanger have been included. There are solutions for overcoming the effects of lighting and occlusion, but they are usually developed for specific groups of objects [12,23] or noises [24].…”