Abstract:Growth stage information of field crops is not only an important basic data for analyzing the relationship between the crop growth process and the agrometeorological conditions, but it is also useful for various aspects of precision agriculture. Up to now, it is primarily obtained manually, which is time-consuming, labor-intensive, subjective and discontinuous. Therefore, a noninvasive method to note observations that also proves to be more efficient, continuous, and automatic is needed. At present, an alternative method based on computer vision has been widely used for monitoring crop growth status due to advantages linked to its low-cost, its intuitiveness and non-contact manner of data gathering it provides. However, little research has been done to improve close observation of different growth stages of field crops using digital cameras. To overcome the drawbacks caused by the current manual observation, a study was conducted to explore the application of computer vision technology for the automatic detection technology of two critical growth stages of maize (emergence and three-leaf stage). In order to identify the growth stages, the first task is to extract the plants from images properly. According to complex factors on farm fields, we proposed a novel crops segmentation method (AP-HI) which is robust and not sensitive to the challenging variation of outdoor luminosity and complex environmental elements. It has laid the foundation for subsequent studies. By virtue of the AP-HI, two automatic detection methods based on imaging were investigated for the two critical growth stages of maize. The former method uses the spatial distribution feature to judge accurately whether the field crop has reached the emergence stage or not. The latter uses the skeleton endpoint to characterize the leaf of seedling and transforms a matter of judgment into that of probability estimation, which leads to the final conclusion. In order to verify the feasibility and validity of our proposed methods, the comparing experiments have been carried out. Five well-established algorithms were utilized to make comparison with AP-HI and its results showed that our method outperformed the other algorithms in yielding the highest performance of 96.68% with the lowest standard deviation of 237%. As for the two automatic detection methods, the crops of two experimental fields located in Zhengzhou, Henan and Taian, Shandong provinces in China were observed both with a human observer and by using automated routines to process images obtained from a camera. In determining the time at which a growth stage occurred, the proposed methods produced the similar results to the manual observation method. Overall, the automated methods can meet the demand for practical observation needed for agronomic modeling and in triggering action alerts to farmers. (C) 2013 Elsevier B.V. All rights reserved.

An Image-Based Approach for Automatic Detecting Tasseling Stage of Maize Using Spatio-Temporal Saliency

In-field automatic detection of maize tassels using computer vision

Maize tassel detection with CA-YOLO for UAV images in complex field environments

YOLOv5-T: A Precise Real-Time Detection Method for Maize Tassels Based on UAV Low Altitude Remote Sensing Images

An Improved Deep Learning Approach for Detection of Maize Tassels Using UAV-based RGB Images

Corn Tassel Detection Based On Image Processing

Detection and Identification of Tassel States at Different Maize Tasseling Stages Using UAV Imagery and Deep Learning

Tassel-YOLO: A New High-Precision and Real-Time Method for Maize Tassel Detection and Counting Based on UAV Aerial Images

Detection of Maize Tassels for UAV Remote Sensing Image with an Improved YOLOX Model

Maize Tassel Detection From UAV Imagery Using Deep Learning

Automatic Detection of Maize Tassels from UAV Images by Combining Random Forest Classifier and VGG16

Maize Tassel Number and Tasseling Stage Monitoring Based on Near-Ground and UAV RGB Images by Improved YoloV8

Tasseled Crop Rows Detection Based on Micro-Region of Interest and Logarithmic Transformation

Maize Tassel Area Dynamic Monitoring Based on Near-Ground and UAV RGB Images by U-Net Model

Multiscale Maize Tassel Identification Based on Improved RetinaNet Model and UAV Images

Towards Fine-Grained Maize Tassel Flowering Status Recognition: Dataset, Theory and Practice.

Analysis of Key Factors Affecting Maize Tassels Detection and Construction of Shared Dataset Based on UAV

Fine-grained maize tassel trait characterization with multi-view representations

Automated Robust Crop-Row Detection in Maize Fields Based on Position Clustering Algorithm and Shortest Path Method.

Automatic image-based detection technology for two critical growth stages of maize: Emergence and three-leaf stage

ESG-YOLO: A Method for Detecting Male Tassels and Assessing Density of Maize in the Field