Palm oil fresh fruit bunch ripeness grading identification using color features

Fruit recognition is useful for automatic fruit harvesting. Fruit recognition application can reduce or minimize human intervention during fruit harvesting operation. However, in computer vision, fruit recognition is very challenging because of similar shapes, colors and textures among various fruits. Illuminations changes due to weather condition also leads to a challenging task for fruit recognition. Thus, this paper tends to investigate the performance of basic Convolutional Neural Network (CNN), Alexnet and Googlenet in recognizing nine different types of fruits from a publicly available dataset. The experimental results indicate that all these techniques produce excellent recognition accuracy, but basic CNN achieves the fastest recognition result compared with Alexnet and Googlenet.

Section: Introductionmentioning

confidence: 99%

Evaluation of CNN, Alexnet and GoogleNet for Fruit Recognition

Muhammad

Nasir

İbrahim

et al. 2018

“…There are many techniques used in color features and among the popular features are color moment [12], [36], [32] and color histogram [2], [22]. A high number of accuracy has been achieved using color moment on palm oil FFB recognition [36]. Therefore, this research implements color moment for color model analysis.…”

Section: A) Feature Extractionmentioning

confidence: 99%

“…According to [2] most of the palm oil grading focuses on two, three or four stages which is unripe, reddish black as underripe, red as ripe, and reddish orange as overripe. Research by [18], [36], [37] focus on two stages which are ripe and unripe of FFB and achieved high accuracy results. [3], [13], [38], [39] recognize three stages of FFB ripeness which is unripe, ripe and overripe while [2], [40] work on four stages which is unripe, underripe, ripe and overripe.…”

mentioning

confidence: 99%

Evaluation of Color Models for Palm Oil Fresh Fruit Bunch Ripeness Classification

Sabri¹,

İbrahim²,

Isa³

2018

This paper investigates the application of eight color models for automatic palm oil Fresh Fruit Bunch (FFB) ripeness classification with multi-class Support Vector Machine (SVM). Ripeness classification is important during harvesting to ensure that they are harvested during the correct ripe stage for optimum oil production. Since color is a significant indicator for agriculturists to determine the ripeness of FFB, it is critical to determine the right color model. Eight color models have been investigated namely, HSV, I1I2I3, LAB, XYZ, YCbCr, YIQ, YUV and RGB. Color moments were extracted from each of these color models for the classification of four stages of FFB ripeness that are unripe, under-ripe, ripe and over-ripe. A database of five hundred images of palm oil FFB has been constructed and experiments showed that YCbCr and YUV outperform the other color models.

“…Previously, research in object recognition uses handcrafted features such as texture features for fall activity recognition [4] and leaf recognition [5]. Besides that, color features have also been applied for fruit recognition [6] where it involves identifying the significant feature and classifier to obtain good recognition results. However, currently, the object recognition research has progressed to Deep Learning (DL) where no handcrafted feature is required and yet the results produced are excellent.…”

Section: Related Workmentioning

confidence: 99%

Celebrity Face Recognition using Deep Learning

Kasim¹,

Rahman²,

İbrahim³

et al. 2018

Face recognition is one of the well studied problems by researchers in computer visions. Among the challenges of this task are the occurrence of different facial expressions like happy or sad, and different views of the images such as front and side views. This paper experiments a publicly available dataset that consists of 200,000 images of celebrity faces. Deep Learning technique is gaining its popularity in computer vision and this paper applies this technique for face recognition problem. One of the techniques under deep learning is Convolutional Neural Network (CNN). There is also pre-trained CNN models that are AlexNet and GoogLeNet, which produce excellent accuracy results. The experimental results indicate that AlexNet is better than basic CNN and GoogLeNet for face recognition.