Unsupervised classification of very high remotely sensed images for grapevine rows detection

Abstract

In viticulture, knowledge of vineyard vigour represents a useful tool for management. Over large areas, the grapevine vigour is mapped by remote sensing usually with vegetation indices like NDVI. To achieve good correlations between NDVI and other vine parameters the rows of a vineyard must be previously identified. This paper presents an unsupervised classification method for the identification of grapevine rows. Only the red channel of an RGB aerial image is considered as input data. The image is first masked preserving only the considered vineyard and then pre-processed with a high pass filter. The pixel populations are split in "row" and "inter-row" subset through a Ward's modified technique. The proposed methodology is compared with standard object oriented procedure tested on six vineyards located in Tuscany using as reference manually digitalized vine rows.

Detection and Segmentation of Vine Canopy in Ultra-High Spatial Resolution RGB Imagery Obtained from Unmanned Aerial Vehicle (UAV): A Case Study in a Commercial Vineyard

“…Another technique used to segment vineyards is the texture analysis method using Fast Fourier Transform (FFT) or the Gabor filters [17,19,20]. However, texture analysis only gives a high performance when vine rows are continuous: the performance decreases when the periodic pattern of the rows is disrupted by row discontinuities caused by missing vines and other vineyard structures (e.g., sheds, irrigation infrastructure, and native vegetation) [18,21]. Therefore, the objective of this study is to compare the performance of four classification methods (K-means, Spectral Indices (SI), Artificial Neural Networks (ANN), and Random Forest (RForest)), for vine canopy detection using ultra-high resolution RGB Imagery acquired with a conventional camera mounted on a low-cost UAV.…”

Individual plant definition and missing plant characterization in vineyards from high-resolution UAV imagery

“…In these systems, despite the high spatial resolution of the sensors currently employed, the outcoming information, such as the vigour zoning, only accounts for averaged data neglecting the contribution of single vine (Arnó, Martínez Casasnovas, Ribes Dasi, & Rosell, 2009). While row detection techniques saw a great development in these last few years (Comba, Gay, Primicerio, & Aimonino, 2015;Delenne, Durrieu, Rabatel, & Deshayes, 2010;Puletti, Perria, & Storchi, 2014;Smit, Sithole, & Strever, 2010), a methodology for single plant detection is still not available. Instead, the ability to recognize automatically single vine within a training row could remarkably improve the representation of the contribution of single plant to the canopy curtain, enabling to detect specific plant pathologies in the row and improving the accuracy of vigour zoning (Lee et al, 2010;Naidu, Perry, Pierce, & Mekuria, 2009;Sankaran, Mishra, Ehsani, & Davis, 2010).…”

Detection of Rows in Agricultural Crop Images Acquired by Remote Sensing from a UAV

“…The neighboring pixel is always coordinate from the centre of the pixel in the image. The region growing of an image A is carried out as shown in equation (5). Where T i is the connected region, i = 1, 2, 3…n and T is the region.…”

“…Row crop detection is fundamental to determining canopy, crop vigor and crop density. Rows detection helps in detection of crop specific regions apriori to provide more reliable correlation of NDVI and other parameters [3]. It also finds application in autonomous path planning and navigation of ground robots.…”

“…Comba [4] described a new algorithm for detection of grape vine rows based on dynamic segmentation, Hough space clustering and total least squares to automatically detect the grape vine rows. Puletti [5] used unsupervised classification for grape vine row detection. The red channel of RGB image was pre processed using a high pass filter and was later analyzed using Ward's modified technique.…”