Issue link: http://winesandvines.uberflip.com/i/987926
June 2018 WINES&VINES 71 WINE EAST GRAPEGROWING - B E C O PA D - Y E A S T & E N Z Y M E S - C R U S H PA D E Q U I P M E N T - S T E R I L E F I LT R AT I O N - W I N E R Y H O S E - O A K A LT E R N AT I V E S EASTERN WINE LABS Serving the Analytical needs of East Coast Wineries WWW.EASTERNWINELABS.COM Ph 609-859-4302 Cell 609-668-2854 chemist@easternwinelabs.com AOAC Member EasternWineLab_Mar09.qxp 1/22/09 9:47 AM Page 1 Basic Hoe comes with a Hillup and a Takeaway Blade. Additional attachments include .3 Tooth Cultivator, Undercutter Blade, Rotary head, " NEW " Rolling Cultivator and "Vine Auger". The Green Hoe Company, Inc. 6645 West Main Road, Portland, NY 14769 PHONE (716) 792-9433 FAX (716) 792-9434 WWW.GREENHOECOMPANY.COM GREEN GRAPE HOE resulting maps are more detailed and useful than those compiled using conventional aircraft. The sensors (cameras) can gather ther- mal, visible, hyperspectral and/or multispectral images for process- ing to obtain data on various canopy variables. A multi-camera array (MCA) is commonly used to obtain raw images needed to produce multispectral images. 2 Depending on the number of sen- sors, several specific light wave- length bands can be measured. An MCA is generally used to measure visible to NIR wavelengths, which allows for measurement of re- flected light from the canopy in those spectral regions. 6 A hyperspectral sensor can measure a much broader range of wavelengths than the 3 to 6 wavelengths measurable with a standard MCA. 36 For example, a hyperspectral sensor could have a range of 260 bands between 400- 885 nanometers (one-billionth of a meter). 37 Drones also can utilize thermal cameras to gather image data to be used to determine can- opy variables such as water stress. 29 An example of thermal camera had spectral range of 7.5- 13 micrometers (one-millionth of a meter) and a -45° to 120° C dy- namic temperature range. 4 As sensor technology has evolved, so has the sensor's resolu- tion of images. The resolution de- pends on the elevation of the UAV over its respective target; there- fore, resolution can range depend- ing on the aims of the study or application. For example, the reso- lution of an MCA can be as low as 0.056 m/pixel at an elevation of 150 m above the study vineyard, 24 a thermal camera can have a reso- lution as low as 0.13 m/pixel at 100 m above the study vineyard 29 and the resolution of a hyperspec- tral camera is 0.4 m/pixel at 575 m above ground level. 8 With better resolution, the quality of the acquired data has improved due to the ease of de- termining the actual crop canopy data in the visible or multispectral image as well as the ease of re- moving the soil or cover crop data from the respective visible, ther- mal or multispectral image. Image processing Raw images taken by the thermal, multispectral, visible, or hyper- spectral sensors require processing with help of computer software before being used for research or precision agriculture. Calibration is done first, and recovers focal distance, point coordinates and lens radial distortion of the cam- era. 4 Several images are taken from different locations and orienta- tions, which allow for the calcula- tion of the variables of the camera and the exterior orientation. Then the pictures are ortho- rectified, a process by which the effects of tilt and terrain of the image are removed. This can be done with a computer program and digital contour maps 24 or by a computer program using an in- ertial measuring unit installed on the UAV that is synchronized with the respective imager. 36 The raw images also require georeferenc- ing, which can be accomplished by using Ground Control Points (GCP) that have been measured with a survey grade GPS device, 33 or by using the data acquired by the UAV's onboard GPS device. Errors in the accuracy of the data point locations are larger without use of GCPs. Once the images are orthorectified and georeferenced, the images can be mosaicked (stitched) together using georef- erenced-based stitching soft- ware. 33 An example of a stitched image is shown in Figure 2a. As UAV platform use is increas- ing, single computer programs are The images acquired by different sen- sors (e.g., thermal, multispectral, visible, hyperspectral) commonly used with UAVs also must be calibrated to avoid possible errors in the data.