The study proposes a novel time‐frequency analysis technique, titled time‐frequency synchroextracting transform which enhances the accuracy of synchroextracting transform. The synchroextracting transform is a powerful postprocessing method that sharpens the energy distribution of spectrogram. However, its limitation is the sole use of time or frequency factors that cannot accurately portray signals containing both ‘slow‐varying’ and ‘fast‐changing’ components. The proposed method applies the synchroextracting procedure in both time and frequency dimensions to improve the energy concentration. A fast Fourier transform‐‐based algorithm is also provided for efficient implementation. The proposed method is validated via numerical and empirical data. Compared with advanced methods including synchrosqueezing transform and synchroextracting transform, the proposed method is suitable to analyse wider range of signals and has better noise robustness than the original synchroextracting transform.

The classification of plug seedlings is important work in the replanting process. This paper proposed a classification method for plug seedlings based on transfer learning. Firstly, by extracting and graying the interest region of the original image acquired, a regional grayscale cumulative distribution curve is obtained. Calculating the number of peak points of the curve to identify the plug tray specification is then done. Secondly, the transfer learning method based on convolutional neural network is used to construct the classification model of plug seedlings. According to the growth characteristics of the seedlings, 2286 seedlings samples were collected to train the model at the two-leaf and one-heart stages. Finally, the image of the interest region is divided into cell images according to the specification of the plug tray, and the cell images are put into the classification model, thereby classifying the qualified seedling, the unqualified seedling and the lack of seedling. After testing, the identification method of the tray specification has an average accuracy of 100% for the three specifications (50 cells, 72 cells, 105 cells) of the 20-day and 25-day pepper seedlings. Seedling classification models based on the transfer learning method of four different convolutional neural networks (Alexnet, Inception-v3, Resnet-18, VGG16) are constructed and tested. The classification accuracy of the VGG16-based classification model is the best, which is 95.50%, the Alexnet-based classification model has the shortest training time, which is 6 min and 8 s. This research has certain theoretical reference significance for intelligent replanting classification work.

During the process of plant protection in agriculture, the distribution and deposition of droplets or fog fields could directly influence the effectiveness and efficiency of spray. The traditional method of measurement of the distribution of droplets mainly used water sensitive papers, glass containers or flour to collect data and inverse results, while a new method of measurement based on the principle of reflection of LIDAR was presented. Droplets were the major targets of the study, and four important algorithms were primarily developed, including the recognition and extraction of targets, the superposition in time-domain, the calculation of effective ranges of distribution, and the development of 3D distribution models. Combined with these algorithms, in order to eliminate the environmental noise, the methods of Fuzzy Environment Matching and Secondary Filter were created and utilized. Meanwhile, the statistics was used for analysis of the duration of scanning as well as computation of the distribution, with enough datasets but the minimum length of time. The results of the experiments showed that the relative error of measurement was less than 7% and Relative Standard Deviation was less than 16%, compared with the values of manual measurement. Furthermore, the 3D models were accurate and clarified in the wind-tunnel experiment. The completed system based on this method could adapt to the requirements of both indoor and outdoor detection. Besides, it is capable of the quantized detection of droplet distribution, providing an effective way of tests for spray technique, especially for the research of the application of plant protection by UAVs. Keywords: droplets, distribution, detection, agricultural aviation, UAV DOI: 10. 25165/j.ijabe.20171004.3118 Citation: Zheng Y J, Yang S H, Lan Y B, Hoffmann C, Zhao C J, Chen L P, et al. A novel detection method of spray droplet distribution based on LIDARs. Int J Agric & Biol Eng, 2017; 10(4): 54–65.

Speed estimation is crucial to monitor the conditions of rotational machinery. Most speed measurements are carried out by installing encoders or tachometers inside the machines. In many cases, such method could be cumbersome or even inaccessible. This paper proposes a vibration-based speed estimation method. The vibration sensors are often cheaper and easier to install than angle encoders. In the proposed method, the continuous wavelet transform (CWT) is used as a preprocessing technique to extract the signal of importance. Then, the time-varying autoregressive (TAR) model is applied to analyze the rotational frequency. Additionally, the paper presents a fast algorithm for implementation. The proposed method is validated by both synthetic and empirical data.