David C. Slaughter

Environmentally adaptive segmentation algorithm for outdoor image segmentation

An environmentally adaptive segmentation algorithm (EASA) was developed for outdoor field plant detection. Based on a partially supervised learning process, the algorithm can learn from environmental conditions in outdoor agricultural fields and build an image segmentation look-up table on-the-fly. Experiments showed that the algorithm can adapt to most daytime conditions in outdoor fields, such as changes in light source temperature and soil type. When compared to a static segmentation technique which was trained under sunny conditions, the EASA improved the image segmentation by correctly classifying 26.9 and 54.3% more object pixels under partially cloudy and overcast conditions, respectively. The improved image segmentation of the EASA technique also allowed up to 32 times more plant cotyledons to be recognized (by leaf morphology) under overcast lighting conditions when compared with a static segmentation technique trained under sunny conditions.

A NIR Technique for Rapid Determination of Soil Mineral Nitrogen

The objective of this investigation is to determine the possibility of rapidly sensing soil mineral-N content using near infrared (NIR) reflectance. Simulation studies were conducted to determine the ability of Partial Least Squares (PLS) and Principal Components Regression (PCR) techniques to relate NIR spectral data to soil nitrate content in the presence of interfering effects and experimental noise. The simulation studies revealed that both PLS and PCR techniques were quite robust in predicting soil nitrate content provided the calibration set included the same interfering effects. These techniques failed completely if the prediction set contained interfering effects which were not included in the calibration set. This implies that a site-specific calibration is necessary for this technique to work successfully. Laboratory tests using Yolo loam and Capay clay soil samples as well as verification tests using field soils (Yolo loam and Capay clay) mixed with nitrogen fertilizer indicated that soil mineral-N content can be determined using the NIR technique provided site-specific calibration is used.

Nondestructive determination of internal quality in peaches and nectarines

A nondestructive optical method for determining the internal quality of intact peaches and nectarines was investigated. The method, based upon visible and near-infrared spectrophotometric techniques, was capable of simultaneously predicting the soluble solids content (r = 0.92), sucrose content (r = 0.87), sorbitol content (r = 0.88), and chlorophyll A content (r = 0.97) of intact peaches and nectarines, and required no sample preparation.

Discriminating Fruit for Robotic Harvest Using Color in Natural Outdoor Scenes

ABSTRACT This research investigated the use of chrominance and intensity information from natural outdoor scenes as a means of guidance for a robotic manipulator in the harvest of fruit. A classification model was developed which could discriminate oranges from the natural background of an orange grove using only color information in a digital color image. A Bayesian classifier correctly classified over 75% of the fruit pixels in the natural scenes analyzed. The decision model was simple enough that a real-time search and centroid calculation technique could be implemented to provide guidance information for the robotic manipulator at the 60 Hz video frame rate.

REAL-TIME X-RAY INSPECTION OF WHEAT FOR INFESTATION BY THE GRANARY WEEVIL, SITOPHILUS GRANARIUS (L.)

A high-resolution real-time x-ray imaging system was assembled using a low-energy, high-current x-ray source, a low-energy x-ray image intensifier, and a CCD camera interfaced to a PC. Overall system resolution was measured at 5 line pairs per mm (100 .m), sufficient for identifying infestations of the granary weevil, Sitophilus granarius (L.) in kernels of wheat. The field of view for imaging of the system was 6 cm2, large enough to image approximately 350 kernels of grain in a single frame. The exposure time for a single frame was 149 ms, yielding a maximum potential throughput rate of around 2500 kernels per second. For this study, 1500 wheat kernels were x-rayed, both on film and with the system described above. Of the imaged kernels, 682 contained infestations ranging in maturity from the egg to the adult life stage of the granary weevil, while 818 were uninfested. Both film and digital images were presented to human subjects to compare recognition of the infested kernels. Overall recognition results averaged 84.4% for the images from the intensifier system vs. 90.2% for the film observations. However, when considering only infestations more advanced than the 3rd larval instar, errors for both sets of images fell below 2% and were not significantly different.

Color Vision in Robotic Fruit Harvesting

ABSTRACT THIS report discusses the use of chrominance information in natural scenes to enhance digital color images used to control a robotic manipulator for fruit harvest. By exploiting the natural high contrast in color between fruit and other objects in a color image the ability to differentiate citrus fruit from background leaves, soil, and sky is demonstrated.

Precision weed control system for cotton

A real–time robotic weed control system was developed and tested in commercial cotton fields. The precision weed control system was capable of distinguishing grass–like weeds from cotton plants and applying a chemical spray only to targeted weeds while traveling at a continuous speed of 0.45 m/s. The robot consisted of a real–time machine vision system, a controlled illumination chamber, and a precision chemical applicator. In commercial cotton fields, the system correctly sprayed 88.8% of the weeds while correctly identifying and not spraying 78.7% of the cotton plants while traveling at 0.45 m/s.

A fruit-tracking system for robotic harvesting

The fruit-tracking system described in this article estimated in real-time (60 Hz) the size and position of a valid fruit region in color images. This information was used to control the motion of a fruit-picking robot. Statistical classification of color pixels was employed to improve image processing under the variety of imaging conditions encountered in a grove. Manual specification of valid fruit colors was possible through interactive training sessions. Object-oriented aperture control allowed adjustment of image exposure based on the illumination level of a targeted fruit. This feature increased the dynamic illumination range of the tracking system by ignoring excessively dark or bright background conditions. The performance of the fruit-tracking system was evaluated with off-line tests and under actual operating conditions.

Nondestructive determination of total and soluble solids in fresh prune using near infrared spectroscopy

A nondestructive optical method for determining the total solids content (TSC) and soluble solids content (SSC) of fresh whole prune (Prunus domestica L. ‘French’) was investigated. The method, based upon near infrared spectrophotometric techniques, could predict the TSC (r 2 /0. 98, SEP/0.80% FW) and SSC (r 2 /0. 96, SEP/ 1.02%) of prunes. A low cost (

Vibration of fresh fruits and vegetables during refrigerated truck transport

2500 2002 USD) diode array-type spectrophotometer was used in the study. # 2002 Elsevier Science B.V. All rights reserved.