Improved deep belief networks and multi-feature fusion for leaf identification

Liu, Nian; Kan, Jiangming

doi:10.1016/j.neucom.2016.08.005

Cited by 50 publications

(31 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…x' adalah nilai desimal yang merupakan konversi dari barisan bilagan biner berdasarkan tetangga pixel x, y adalah pixel tetangga x dan N adalah jumlah pixel tetangga x. Setelah semua pixel pada citra gray level diinisialisasi dengan nilai baru, Fitur LBP didapatkan dari nilai histogram pada citra gray level yang telah ditransformasi ke dalam bentuk 1dimensi (Liu & Kan, 2016).…”

Section: Nounclassified

See 1 more Smart Citation

Klasifikasi Tipe Sel Normal/Abnormal Berdasarkan Citra Pap-Smear Menggunakan Convolutional Neural Network

Fathurahman

Nurmufti

Suherlan

2019

JTIY

View full text Add to dashboard Cite

The classification of cell types plays an essential role in monitoring the growth of cancer cells. One of the methods to determine the cancer type is to analyze the pap-smear images manually. Nevertheless, the manual analysis of pap-smear images by the expert has several limitations, such as time-consuming and prone to misdiagnosis. For reducing the risks, it requires the automatic classification of cell types based on pap-smear images. This study utilizes the convolutional neural network (CNN) architectures to automatically classify the cell type into two-class categories (normal/abnormal) based on three features. These features, such as the local binary pattern, gray level co-occurrence matrix, and shape features, are extracted from pap-smear images. This study shows the performance of CNN achieved the maximum accuracy of 99.98%, 100.0%, 99.78% in training, validation, and testing data. Our approach also outperforms the performance of the baseline methods. Keywords : CNN, Classification, Cell, Neural Network, Pap-smear

show abstract

Section: Nounclassified

“…Gray level co-occurence matrix (GLCM) adalah metode ekstrasi fitur berdasarkan matrik tabulasi dari setiap nilai pixel pada citra gray scale (Liu & Kan, 2016). Nilai fitur yang dapat diperoleh dari GLCM adalah contrast, homogeneity, energy dan correlation.…”

Section: Gray Level Co-occurence Matrix (Glcm)unclassified

Klasifikasi Tipe Sel Normal/Abnormal Berdasarkan Citra Pap-Smear Menggunakan Convolutional Neural Network

Fathurahman

Nurmufti

Suherlan

2019

JTIY

View full text Add to dashboard Cite

show abstract

“…Other researches had used a combination of geometric and textural, allowing them to use dried, wet or even misshapen leaves [12]. Some author combine both textural information and shape features to identify leaves [13,14].…”

Section: State Of the Artmentioning

confidence: 99%

“…However, a recurring problem in the leaves of plants is that the chromaticity in the leaves is not static, it is variable with respect to time and commonly with respect to other factors. Other authors consider in addition to chromaticity and form, the texture of the leaf [16] or use combinations of descriptors to improve the classification performance [13,14].…”

Section: State Of the Artmentioning

confidence: 99%

Complex Identification of Plants from Leaves

Cervantes

García-Lamont

Rodríguez-Mazahua

et al. 2018

Intelligent Computing Methodologies

View full text Add to dashboard Cite

The automatic identification of plant leaves is a very important current topic of research in vision systems. Several researchers have tried to solve the problem of identification from plant leaves proposing various techniques. The proposed techniques in the literature have obtained excellent results on data sets where the leaves have dissimilar features to each other. However, in cases where the leaves are very similar to each other, the classification accuracy falls significantly. In this paper, we proposed a system to deal with the performance problem of machine learning algorithms where the leaves are very similar. The results obtained show that combination of different features and features selection process can improve the classification accuracy.

show abstract

“…The extracted shape descriptor is invariant to rotation, translation, scale, and start point of the edge centroid distance sequences. Liu and Kan [ 10 ] extracted a lot of texture and shape features for leaf classification, including Gabor filters, local binary patterns, Hu moment invariants, gray level cooccurrence matrices, and Fourier descriptors and applied deep belief networks with dropout to classify plant species. Vijayalakshmi and Mohan [ 11 ] extracted gray level cooccurrence matrix (GLCM) and local binary pattern (LBP) features for leaf classification.…”

Section: Introductionmentioning

confidence: 99%

Discriminant WSRC for Large-Scale Plant Species Recognition

Zhang

Zhu

et al. 2017

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

In sparse representation based classification (SRC) and weighted SRC (WSRC), it is time-consuming to solve the global sparse representation problem. A discriminant WSRC (DWSRC) is proposed for large-scale plant species recognition, including two stages. Firstly, several subdictionaries are constructed by dividing the dataset into several similar classes, and a subdictionary is chosen by the maximum similarity between the test sample and the typical sample of each similar class. Secondly, the weighted sparse representation of the test image is calculated with respect to the chosen subdictionary, and then the leaf category is assigned through the minimum reconstruction error. Different from the traditional SRC and its improved approaches, we sparsely represent the test sample on a subdictionary whose base elements are the training samples of the selected similar class, instead of using the generic overcomplete dictionary on the entire training samples. Thus, the complexity to solving the sparse representation problem is reduced. Moreover, DWSRC is adapted to newly added leaf species without rebuilding the dictionary. Experimental results on the ICL plant leaf database show that the method has low computational complexity and high recognition rate and can be clearly interpreted.

show abstract

Improved deep belief networks and multi-feature fusion for leaf identification

Cited by 50 publications

References 25 publications

Klasifikasi Tipe Sel Normal/Abnormal Berdasarkan Citra Pap-Smear Menggunakan Convolutional Neural Network

Klasifikasi Tipe Sel Normal/Abnormal Berdasarkan Citra Pap-Smear Menggunakan Convolutional Neural Network

Complex Identification of Plants from Leaves

Discriminant WSRC for Large-Scale Plant Species Recognition

Contact Info

Product

Resources

About