English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

With the development of information technology, precision agriculture has also ushered in new development prospects. The use of farm robots to accurately identify the navigation path is of great significance for achieving accurate positioning of agriculture. In this study, when analysing the extraction algorithm of farm robot visual navigation based on dark primary colours, a method for pre-processing and edge detection of farmland images based on dark primary colours is proposed. At the same time, the least square method of linear fitting is used for the navigation path of agricultural robot, and then the fitting program is executed. On this basis, combined with the actual situation of outdoor farming and greenhouse cultivation of crops, the effectiveness of the robotic visual navigation extraction algorithm was verified. The research results show that for any form of farmland cultivation, image extraction technology based on dark primary colours can effectively distinguish between soil and crops, and the visual navigation path of farm robots fitted with least squares is basically linear, which is consistent with the commonly used crops for farm planting. The legal route is basically the same, and then the effectiveness of the extraction algorithm is verified. It is hoped that this study will provide a certain reference and reference for the analysis of the field navigation robot visual navigation extraction algorithm based on dark primary colours. 摘要 随着信息技术的发展,精准农业也迎来了新的发展前景。使用农业机器人准确识别导航路径对实现农业 的精确定位具有重要意义。本文在分析基于深色原色的农用机器人视觉导航提取算法的基础上,提出了一种基 于深色原色的农田图像预处理和边缘检测方法。同时,线性拟合的最小二乘方法用于农用机器人的导航路径。 执行拟合。在此基础上,结合野外耕作和温室大棚栽培的实际情况,验证了机器人视觉导航提取算法的有效 性。研究结果表明,对于任何形式的农田耕作,基于深色原色的图像提取技术都可以有效地区分土壤和农作 物,配备最小二乘法的农场机器人的视觉导航路径基本上是线性的,这与通常的做法是一致的。用于农作物种 植的农作物。合法路径基本相同,然后验证了提取算法的有效性。希望这项研究为基于深色原色的野外导航机 器人视觉导航提取算法的分析提供一定的参考和借鉴。 INTRODUCTION Agriculture is the basis for human survival, an important guarantee for human food and clothing, and plays an important role in economic prosperity and social stability. With the development of social economy, the development of agricultural modernization is steadily advancing. One of the most important signs of agricultural modernization is the increasing degree of agricultural mechanization (Teshome and Degu, 2019). With the rapid development of information technology, agricultural mechanization is moving to the next new development stage, namely, agricultural precision. The so-called precision agriculture refers to relying on information technology and computer technology such as artificial intelligence to achieve the precise development of agriculture. Among them, farmland cultivation is the most important and basic part of agriculture, and is a part of the vigorous development of precision agriculture (Srbinovska and al., 2015). Machine vision has been widely used in the development of precision farming technology. Field robot is of great significance for accurate identification of navigation path and accurate positioning. 1 As. PhD. Lec. Zhongkun Hou Vol. 62, No. 3 / 2020 INMATEH – 220 Machine vision refers to the use of cameras and other sensors to collect images, and to simulate human visual functions through image processors. After completing a series of image acquisition and simulation function operations, extract and understand relevant information from the images, and finally use the relevant information for monitoring and control (Sun T.H., Tien F.C. and Tien F.C., 2016). Farm robot visual navigation is relatively low in cost and easy to operate. It has no special requirements for operators and agricultural scale, and has good applicability, so it is widely used in farm cultivation. In addition, in the visual navigation of farm robots, the extraction of navigation lines is the most critical content, which is the key to determine the navigation accuracy and speed of farm robots. In this study, when analysing the field navigation robot visual navigation extraction algorithm based on dark primary colours, a method for pre-processing and edge detection of farmland images based on dark primary colours is proposed. At the same time, the least square method in linear fitting is used to extract the navigation lines. Farm robot navigation path to fit. On this basis, combined with the actual planting conditions of outdoor farmland cultivation and greenhouse cultivation of crops, the effectiveness of the robot visual navigation extraction algorithm was verified. The innovation of this research lies in the use of dark primary colour prior theory for grayscale preprocessing of farmland images, which can quickly perform grayscale processing on the image, and can effectively distinguish crops and soil in farmland images. At the same time, the research also focused on the practical application of extraction algorithms, combining the method with the actual farmland cultivation, and by combining the actual conditions of outdoor farmland cultivation and greenhouse cultivation of crops, the effectiveness of the robotic visual navigation extraction algorithm was verified. This research mainly includes four parts. The first part is the application of the prior theory of dark primary colours by domestic and foreign scholars and the research of robot visual navigation. The second part is the research of the field navigation robot visual navigation extraction algorithm based on the dark primary grayscale processing, including the use of dark primary colour images for grayscale pre-processing and edge detection, and the least square method in linear fitting to extract the navigation path. The third part is to verify the effectiveness of the robotic visual navigation extraction algorithm in combination with the actual planting conditions of outdoor farmland cultivation and greenhouse cultivation. Aimed at the inaccuracy of the Gaussian algorithm in the dark primary colour prior theories, Li A et al. proposed a method to improve the image clarity based on the dark channel, and verified this by analysing the factors affecting the image colour shift. Regarding the effectiveness of the method, the results show that this method can effectively solve the problem of image migration in light-coloured areas (Li A., Fang Z and Mi B., 2018). In the study of coal dust and smoke images, Mingben D.U et al proposed an image enhancement algorithm based on dark primary color prior theory. Based on the image degradation model, the image denoising technology is realized. The experimental results show that the image denoising technology can optimize and enhance the image quality. (Mingben D.U., et al., 2015). Xiao JS et al proposed an image denoising algorithm based on the dark primary colour prior theory to overcome the colour distortion and halo effect of the sky, accurately estimate the overall atmospheric light in the sky area, and use the median filter technology to obtain detailed Edge information to effectively estimate the transmission image (Xiao J.S. et al., 2017). Liu H.B. et al. proposed an image defogging method based on the a priori theory of dark primary colours, and changing the grey value of the image through linear mapping, thereby improving the medium transmittance in foggy and large sky areas. This method can achieve faster image processing speed (Liu H.B., Yang J. and Wu Z.P., 2015). When studying the single image denoising algorithm, Alharbi EM et al. proposed an atmospheric light estimation algorithm based on the dark primary colour prior theory. By improving the attenuation image contrast, etc., the effectiveness of the algorithm in estimating atmospheric light was verified (Alharbi E.M., Ge P. and Wang H., 2016). When studying the visual navigation of autonomous mobile robots, Kurashiki K. et al. proposed an image-based road boundary tracking control law, and designed an interactive model of robot motion and image features. This method reduces a priori information and improves the map scalability (Kurashiki K, et al., 2015). Rômulo T Rodrigues and others proposed a method of improving robot visual positioning based on artificial potential field when researching robot visual navigation, which connects the features in robot visual images with potential energy to promote the robot to advance. The research results show that this method has good performance (Rômulo T. Rodrigues, et al., 2018). Phalak Y. et al., when studying robot navigation, combined with the parabolic geometry in mathematical tools, proposed a method of navigation using image processing, and finally proposed a system expansion based on the actual experimental results Vol. 62, No. 3 / 2020 INMATEH – 221 of the combination of static and dynamic (Phalak Y. et al., 2018). Kamaev A.N. et al. used the point cloud computing algorithm and the algorithm to establish the point cloud correspondence to provide navigation data for the aircraft when studying the navigation problem of the automatic launch vehicle. The test results show that this method can make the aircraft always follow the navigation trajectory (Kamaev A.N. and Karmanov D.A., 2018). When studying the visual navigation path of coloured cotton robots, Li D. et al. proposed a field navigation path extraction method based on horizontal spline segmentation, and obtained the connected domain of cotton valley by detecting the connected domain in the horizontal spline. Planning the navigation path for the coloured cotton robot finally verified the accuracy and effectiveness of the method (Li D., Xu S. and Zheng Y., 2017). Kumar B. et al. proposed a robot path navigation method using RF communication and wireless audio and video transmission when studying the mobile navigation of autonomous robots. By installing a navigation embedded system inside the robot, the robot can help the robot to achieve autonomous mobile navigation. The actual test verifies the effectiveness of the method (Yang H., Gao L. and Tang N., 2019). Delaune J. et al propose

ANALYSIS OF VISUAL NAVIGATION EXTRACTION ALGORITHM OF FARM ROBOT BASED ON DARK PRIMARY COLOUR