Automatic detection of bunches of grapes in natural environment from color images

Reis, Manuel J. C. S.; Morais, Raul; Peres, Emanuel; Pereira, Carlos; Contente, Olga; Soares, Salviano; Valente, António; Baptista, José; Ferreira, Paulo J. S. G.; Cruz, José Bulas

doi:10.1016/j.jal.2012.07.004

Cited by 97 publications

(62 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Other studies propose the use of alternative sources of information to facilitate or automate bunch weight estimates. This is the case of image analysis that has been proposed to detect, count and estimate the weight of clusters (Diago et al 2012;Reis et al 2012;Nuske et al 2011;Serrano et al 2005;Dunn and Martin 2004) or to estimate the number and the volume of berries (Grocholsky et al 2011;Rabatel and Guizard 2007). Other authors have proposed continuous weighing devices positioned on the trellising system (Blom and Tarara 2009) at a specific location in the field.…”

Section: Introductionmentioning

confidence: 99%

Use of multi-spectral airborne imagery to improve yield sampling in viticulture

et al. 2015

View full text Add to dashboard Cite

The wine industry needs to know the yield of each vine field precisely to optimize quality management and limit the costs of harvest operations. Yield estimation is usually based on random vine sampling. The resulting estimations are often not precise enough because of the high variability within vineyard fields. The aim of the work was to study the relevance of using NDVI-based sampling strategies to improve estimation of mean field yield. The study was conducted in nine non-irrigated vine fields located in southern France. For each field, NDVI was derived from multi-spectral airborne images. The variables which define the yield: [berry weight at harvest (BWh), bunch number per vine (BuN) and berry number per bunch (BN)] were measured on a regular grid. This database allowed for five different sampling schemes to be tested. These sampling methods were mainly based on a stratification of NDVI values, they differed in the way as to whether NDVI was used as ancillary information to design a sampling strategy for BuN, BN, BW or for all yield variables together. Results showed a significant linear relationship between NDVI and BW, indicating the interest of using NDVI information to optimize sampling for this parameter. However this result is mitigated by the low incidence of BW in the yield variance (4 %) within the field. Other yield components, BuN and BN explain a higher percentage of yield variance (60 and 11 % respectively) but did not show any clear relationship with NDVI. A large difference was observed between fields, which justifies testing the optimized sampling methods on all of them and for all yield variables. On average, sampling methods based on NDVI systematically improved vine field yield estimates by at least 5-7 % compared to the random method. Depending on the fields, error improvement ranged from -2 to 15 %. Based on these results, the practical & B. Tisseyre recommendation is to consider a two-step sampling method where BuN is randomly sampled and BW is sampled according to the NDVI values.

show abstract

Section: Introductionmentioning

confidence: 99%

Use of multi-spectral airborne imagery to improve yield sampling in viticulture

et al. 2015

View full text Add to dashboard Cite

show abstract

“…The OHTA color space was transform from the RGB color model through linear conversion and the extraction accuracy was more than 95%. When the fruits and leaves have similar colors, color-based segmentation methods are not available for recognizing fruits (Reis et al, 2012).…”

Section: Single Feature Analysis Approachesmentioning

confidence: 99%

A review of key techniques of vision-based control for harvesting robot

Zhao

Gong

Huang

et al. 2016

Computers and Electronics in Agriculture

303

126

View full text Add to dashboard Cite

“…Métodos que se han basados principalmente en LIDAR (siglas en inglés de Light Detection and Ranging) (Fleck et al, 2004;Omasa et al, 2007;Garrido et al, 2012;Arnó et al, 2013;Correa et al, 2013b) y en procesamiento de imágenes en el espectro visible (Dunn & Martin, 2004;Rabatel et al, 2007;Wang et al, 2012;Reis et al, 2012), imágenes que, en términos generales, son tomadas con cámaras digitales convencionales.…”

Section: Sensores Y Tecnologías De Imagenunclassified

“…Además, a través de éstas imágenes es posible estimar y proyectar rendimientos con base la uva visible en la imagen (Dujnn & Martin, 2004;Diago et al, 2013) Estos sistemas de caracterización generalmente no operan en tiempo real y los que sí lo hacen están diseñados para identificar sólo una característica a la vez. Por ejemplo, Reis et al (2012) detecta uvas, en imágenes tomadas durante la noche, usando un algoritmo que establece rangos de color para la uva, rangos que no son aplicables a otros viñedos ya que son producto de prueba y error para un viñedo en particular y en un estado de desarrollo particular. Para otro frutal, cerezos en particular y tendiente a detectar ramas individuales, se ha recurrido a la misma estrategia de capturar imágenes durante la noche y con iluminación artificial (Amatya et al, 2016) para evitar los problemas ocasionados por la variabilidad de las condiciones de iluminación dentro de la escena.…”

Section: Sensores Y Tecnologías De Imagenunclassified

“…Por ejemplo Reis et al (2012) alcanzan un 97 % de aciertos en uvas rojas, adquiriendo las imágenes durante la noche y fotografiando directamente el racimo como se puede ver en la Figura 50 d). Debe notarse que Liu et al (2013) informan que la técnica propuesta por Reis et al (2012) Otro ejemplo lo encontramos en Casser, (2016) quien también emplea imágenes nocturnas, las que clasifica empleando SVM sobre el espacio de color L*a*b* con resultados de 99 % pero solo aplicadas a un conjunto de 12 imágenes. Este pequeño conjunto no permite garantizar capacidad de generalización.…”

Section: Cámara De Profundidadunclassified

See 1 more Smart Citation

Extracción de Características en Viñedos Mediante Algoritmos de Clasificación Difusa y Técnicas de Inteligencia Artificial

Farias¹

View full text Add to dashboard Cite

show abstract

Automatic detection of bunches of grapes in natural environment from color images

Cited by 97 publications

References 21 publications

Use of multi-spectral airborne imagery to improve yield sampling in viticulture

Use of multi-spectral airborne imagery to improve yield sampling in viticulture

A review of key techniques of vision-based control for harvesting robot

Extracción de Características en Viñedos Mediante Algoritmos de Clasificación Difusa y Técnicas de Inteligencia Artificial

Contact Info

Product

Resources

About