Search

High-clearance plant protection machine plays an important role in the large-scale chemical fertilizer spraying machinery in China. The wheels of high-clearance plant protection machine must be driven along the ridge during the working process. However, the driver's vision is blocked by corn plants, which is easy to cause corn seedlings damaged by tires. In view of the serious phenomenon of seedling injury during the operation of high-clearance plant protection machine, a method of real-time extraction of navigation line between the corn rows based on a dynamic Region of Interest (ROI) was proposed in this paper. According to the different growing periods of corn, the videos were collected when the average height of corn plant was 30, 50 and 70 cm under different light conditions. Through reducing the influence of light by pixel normalization, segmenting the image of corn and background by green image processing algorithm and the maximum class square error method, the images were enhanced and removed by morphological processing. Firstly, the collected image was divided into 10 image bands, and each image band was processed by vertical projection method to locate the center lines between crop rows in the first frame of the video. The approximate position of the static ROI was determined and marked based on the perspective principle of camera and geometric information of the corn seedling belts. It took 473.7 ms to divide the static ROI, which does not affect the real-time performance of navigation since the determined of static ROI was prior to the navigation operation. Secondly, the static ROI was applied in the second frame of video, in which the navigation line between the corn rows was obtained by sub-regional feature points clustering algorithm. According to the geometric information of navigation line and ROI, the ROI was updated and optimized. Then the updated ROI was applied to the navigation line extraction of the next video frame. 225 consecutive videos were selected to divide the ROI manually, and the dynamic ROI obtained by proposed method was compared with the artificial ROI. The pixel deviation between the artificial ROI and the ROI was about 10 pixels, which shown that the algorithm had good convergence and met the requirements of navigation accuracy. Finally, a comparative experiment of artificial navigation line extraction and the algorithm in this paper by randomly selecting 100 frames of image. The results showed that the algorithm proposed by this paper could automatically obtain the dynamic ROI, the average error of navigation line extraction between manual and algorithm in this paper was 1.157°, and the accuracy rate of corn seedling belt recognitionwas 96%, and the average processing time of a real-time image with a resolution of 1 920 pixels×1 080 pixels was 97.56 ms. July 16, 2019, the navigation experiment was carried out in corn field, and the heading angle was used as the input signal of the automatic steering control system. The results showed that the wheels of high-clearance plant protection machine walked along the corn rows, and the wheel angle floated up and down at 0°. The algorithm proposed in this paper can meet the accuracy and real-time requirements of the automatic walking for high-clearance plant protection machine in corn. [ABSTRACT FROM AUTHOR]

Pomegranates are one of the economic fruits in China. The timely detection of pomegranate diseases and the corresponding preventive measures are important to increase crop yields and reduce economic losses. To tackle the issues that traditional edge detection operators usually resulted in low accuracy of the detection of pomegranate diseases and its contour, low anti-noise capability, and created false edges, this study presented an improved Prewitt operator with an adaptive threshold. Firstly, in the image pre-processing stage, the pending image was enhanced by high-frequency emphasize filter, and the high-frequency component, which represented to the details of the pomegranate sample in the image, was increased by the filter. On the contrary, the low-frequency component, which represented to the background of the pomegranate sample, was attenuated to facilitate the following image processing. Since the most original images contained the Gaussian noise, bilateral filtering was used to process the noise present in the image. The weighted average of the brightness values of adjacent pixels was used to represent the intensity of a pixel. The weighted average method was used based on Gaussian distribution. The weight calculation took into account both the Euclidean distance between the pixels and the radiation difference in the pixel range domain to better maintain edge feature information. Secondly, a fifth-order convolutional mask was proposed. The weights of the elements in the mask were set according to the properties of the Gaussian noise probability distribution to reduce the effect of noise on the algorithm. The weight of the central element of the convolutional mask was increased so that the edge information of the image had higher contrast. In terms of gradient calculation, eight direction templates were used to perform the convolutional operation of the image, then the gradient values of each direction were calculated. After that, the corresponding gradient values were obtained, and the gradient values of eight directions were combined into eight groups according to the rule of orthogonality. As a result, there was 90 degrees difference between each group of gradients. Then calculated the L2 norm of each set of gradients, and used the largest value of L2 norm as the gradient of the current pixel. It was shown that it was only necessary to calculate the convolutions in half the direction due to the convolutional results in opposite directions were inverse to each other. The gradients of the rest of the direction could be obtained by its symmetric property. The adaptive threshold was implemented through the minimum error method to decrease the probability of false detection and prune irrelevant details. The mean and variance of the target and background were calculated, meanwhile, the minimum error objective function was obtained according to the principle of the minimum error classification. The value that minimizes the objective function was taken as the optimal threshold. To demonstrate the effectiveness of the method in this paper, 607 test images were manually collected as samples in the natural environment from different angles and different illumination conditions. To further reduce the interference of light changes and suppress background noise, the pending images were converted into grayscale images. Therefore, the algorithm was tested using the converted grayscale images. The experimental results showed that the algorithm proposed in this study achieved a better definition at the edge of the lesion. Compared with the traditional edge detection operator, it achieved a higher recognition accuracy and lower running time. The peak signal to noise ratio of the obtained image was higher, and the obtained edge was complete and accurate. The algorithm proposed in this paper could quickly and accurately detect the area of pomegranate lesions, providing a fundamental reference for the future prevention of pomegranate diseases. [ABSTRACT FROM AUTHOR]

Tomato has become the most cultivated and consumed vegetable crop in the world, and China has become one of the largest producers and consumers of tomatoes in the world. The pectin content in cell wall of tomato during ripening is closely related to fruit hardness, which is one of the important indicators to determine the maturity and reflect the quality of tomato. The requirement of tomato maturity classification and evaluation promotes the development of non-destructive, fast and accurate detection methods of tomato hardness.Hyperspectral imaging integrates spectroscopy and imaging technology in an analysis system, which transfers tomato maturity assessment from subjective, manual classification and measurement methods. Hyperspectral imaging has been widely used in the rapid acquisition of information to classify, detect or identify the quality of various fruits. A novel method for tomato hardness detection based on hyperspectral imaging and modified interval Random Frog was proposed in this paper. Firstly, hyperspectral images of 120 tomato samples in different mature periods were captured by hyperspectral imaging system covering near-infrared region (865.11nm-1 711.71nm). And the hardness data of tomato was obtained by texture analyzer. Secondly, the spectral data were pretreated by multiplicative scatter correction (MSC) and normalized preprocessing to eliminate noise and improve signal-to-noise ratio. The validity of the characteristic wavelength plays a crucial role in the prediction performance of the model. Therefore, we need an effective method to extract the effective wavelength to improve the accuracy of the model. Interval random frog (iRF) algorithm considers all possible spectral wavelengths and ranks all the wavelengths based on selected probability. But one of the disadvantages of this method is large number of iterations and slow convergence. In view of above disadvantages, the traditional iRF algorithm was optimized in terms of constructing initial variable subset method. A modified interval Random Frog (miRF)was proposed to extract the characteristic wavelength effectively. Finally, a prediction model was developed based on partial least squares regression (PLSR) method to detect tomato hardness. The results indicated that the convergence efficiency and accuracy of miRF has a significantly improvement compared with the iRF method. The iRF has selected 100 feature bands, accounting for 40.65% of the full band, and its runtime was 32.1min. miRF has selected 47 feature bands, accounting for 19.1% of the full band, and its runtime was 1.6 min. It can be seen that miRF greatly reduces the running time of the algorithm. The characteristic wavelengths selected by iRF and miRF methods were mainly distributed in 1 582 nm-1 655 nm, followed by 1 160 nm-1 190 nm and 1 353 nm-1 383 nm, indicating that above regions were sensitive bands to tomato hardness. In order to prove the effectiveness of the proposed algorithm, the results of miRF-PLSR were compared with those of iRF-PLSR and SPA-PLSR. The prediction set correlation coefficients (RP) of the SPA-PLSR model and the iRF-PLSR model were 0.803 9 and 0.936 6 respectively. And the RP of miRF-PLSR model was 0.968 5. The root mean square error (RMSEP) of the SPA-PLSR model and the iRF-PLSR model were 0.007 7 kg/mm2 and 0.004 4 kg/mm2respectively. And the RMSEP of miRF-PLSR model was 0.004 0 kg/mm2. The experiments results show that the miRF-PLSR model has the best prediction results in all models. [ABSTRACT FROM AUTHOR]

With the development of Internet, aquaculture information is increasing rapidly in decade years in both internet and the Internet of Things (IoT). Now users are confronted with the information overload, and most of the information has not been effectively utilized. In order to select requested resource for specific users from ubiquitous resource, an aquaculture recommender system was built in this paper. Several IoTs had been deployed in many provinces of China, like Beijing, Jiangsu and Shandong. What's more, an IoT platform was established to collect IoT environment information in real time and crawl aquaculture information from Internet. In this research, item based collaborative filtering algorithm was combined with Web usage mining to generate recommendation. Web usage mining technology processed Web logs in host server and analyzed browsing behavior of users to establish user preference model. Furthermore, item based collaborative filtering used collective intelligence to recommend items that was similar with the items user prefers. The producers of this research were as follows. First, an aquaculture household interest questionnaire was designed after field research with aquaculture households. The questionnaire included user basic investigation, pond basic investigation and user interest investigation. Second, users rated the items in the questionnaire according to their interests. These items included aquaculture technology, fish disease prevention knowledge, government policies, pond production information, disease warning information, weather information, information output prediction and fault information etc. So that user tendency was initialized through the questionnaire and system registration. Third, further user interest could be collected in the form of Web logs. Web logs recorded user browsing behavior, such as user name, IP, access date and time and visited pages. User browsing behavior indicated user interests. To a certain extent, the longer time user browsers a page, the more interest user have. This technology was called Web usage mining. After user initial model and Web usage mining, the user-item rating matrix was calculated. Fourth, item-based collaborative filtering calculated the similarity between two items. The methods of similarity calculation mainly included Pearson similarity, cosine similarity, general modified cosine similarity and improved modified cosine similarity. Considering different user rating behaviors, some users would rate much higher or lower score than the average. So improved modified cosine similarity method was used to reduce the impact of user rating behavior. Using this method, all ratings of one user would be divided by highest rating, and the new user-item rating matrix was obtained. Fifth, the prediction rating from users to items was produced based on item similarity. The recommendation system pushed the items with highest prediction rating to the corresponding users using the top neighborhood method. Finally, this research was evaluated by mean absolute error. Results showed that when recommend items were more than 4, MAE of improved modified cosine method were the least in all methods. So that, the improved modified cosine method (IMCM) was chosen to calculate item similarity. In conclusion, the cold start problem of recommender system could be solved by initial user interest like user registration and questionnaire investigation. And the IMCM method improves the accuracy by reducing impact of user behavior. The aquaculture system recommends precisely according to user interest from both Internet and Internet of Things (IoT), such as trading information, farming technology, government policies and Internet of Things data. There are further works to do in the future. Context aware can be taken into consideration to improve recommendation precision, such as time, location and fault information in pond. Time presents the different production season of aquatic products; location presents where the pond is; and the fault information presents the abnormal status of IoT in pond. [ABSTRACT FROM AUTHOR]