S.M. Mohidul Islam

Work place: Computer Science and Engineering Discipline, Khulna University, Khulna, Bangladesh

E-mail: mohid@cse.ku.ac.bd


Research Interests: Data Mining, Image Processing, Pattern Recognition, Machine Learning


S.M. Mohidul Islam is an Associate Professor at the Computer Science and Engineering Discipline, Khulna University, Bangladesh. He received his B.Sc. Engg. and M.Sc. Engg. degree from Khulna University. His research interests include Machine learning, Data Mining, Pattern Recognition, and Digital Image Processing.

Author Articles
Rice Leaf Disease Recognition using Local Threshold Based Segmentation and Deep CNN

By Anam Islam Redoun Islam S. M. Rafizul Haque S.M. Mohidul Islam Mohammad Ashik Iqbal Khan

DOI: https://doi.org/10.5815/ijisa.2021.05.04, Pub. Date: 8 Oct. 2021

Timely detection of rice diseases can help farmers to take necessary action and thus reducing the yield loss substantially. Automatic recognition of rice diseases from the rice leaf images using computer vision and machine learning can be beneficial over the manual method of disease recognition through visual inspection. During the recent years, deep learning, a very popular and efficient machine learning algorithm, has shown great promise in image classification task. In this paper, a segmentation-based method using deep neural network for classifying rice diseases from leaf images has been proposed. Disease-affected regions of the rice leaves have been segmented using local segmentation method and the Convolutional Neural Network (CNN) has been trained with those images. Proposed method has been applied on three different datasets including the one created by us which consists of the rice leaf images collected from Bangladesh Rice Research Institute (BRRI). Three state-of-the-art CNN architectures VGG, ResNet and DenseNet, used in the proposed method, have been trained with these three datasets for classifying the diseases. Classification performance of the proposed method using the said three CNN architectures for the three datasets have been analyzed and compared. These results show that this model is quite promising in classifying rice leaf diseases. Outcome of this research is an enhancement in the performance of rice disease classification which is quite significant for the viability of this work to be transformed into a real-time application for the farmers.

[...] Read more.
Colour, Texture, and Shape Features based Object Recognition Using Distance Measures

By S.M. Mohidul Islam Farhana Tazmim Pinki

DOI: https://doi.org/10.5815/ijem.2021.04.05, Pub. Date: 8 Aug. 2021

Object recognition is the recognizing process of objects into semantically expressive classes using its visual insides. Classification of objects becomes complex and challenging task because of its size, poor image quality, occlusion, scaling, geometric distortion, lightening, etc. In this paper, global descriptors that means Color, Texture, and Shape features are used to recognize object. Color histogram is used to obtain the color content, texture content is obtained using Gabor wavelet, and shape content is extracted using Hough transform. These low level or global features are used for creating feature vector. Distance measure is used to find the 1-Nearest Neighbor from the training images i.e. object with minimum distance or maximum similarity with visual contents of the query image. The class of that training image is the predicted label of the query image. We have used twelve different distance measures: some are metrics, some are non-metrics and finally, their recognition accuracy is compared. Ensemble of these distance measures is also used for object recognition in the image. We evaluate this method on a publicly available object-recognition dataset: Columbia Object Image Library (COIL-100) dataset. The experiments show that the recognized results outperform many state-of-the-art methods.

[...] Read more.
Content-based Fish Classification Using Combination of Machine Learning Methods

By S.M. Mohidul Islam Suriya Islam Bani Rupa Ghosh

DOI: https://doi.org/10.5815/ijitcs.2021.01.05, Pub. Date: 8 Feb. 2021

Fish species recognition is an increasing demand to the field of fish ecology, fishing industry sector, fisheries survey applications, and other related concerns. Traditionally, concept-based fish specifies identification procedure is used. But it has some limitations. Content-based classification overcomes these problems. In this paper, a content-based fish recognition system based on the fusion of local features and global feature is proposed. For local features extraction from fish image, Local Binary Pattern (LBP), Speeded-Up Robust Feature (SURF), and Scale Invariant Feature Transform (SIFT) are used. To extract global feature from fish image, Color Coherence Vector (CCV) is used. Five popular machine learning models such as: Decision Tree, k-Nearest Neighbor (k-NN), Support Vector Machines (SVM), Naïve Bayes, and Artificial Neural Network (ANN) are used for fish species prediction. Finally, prediction decisions of the above machine learning models are combined to select the final fish class based on majority vote. The experiment is performed on a subset of ‘QUT_fish_data’ dataset containing 256 fish images of 21 classes and the result (accuracy 98.46%) shows that though the proposed method does not outperform all existing fish classification methods but it outperforms many existing methods and so, the method is a competitive alternative in this field.

[...] Read more.
Pedestrian Detection in Thermal Images Using Deep Saliency Map and Instance Segmentation

By A. K. M. Fahim Rahman Mostofa Rakib Raihan S.M. Mohidul Islam

DOI: https://doi.org/10.5815/ijigsp.2021.01.04, Pub. Date: 8 Feb. 2021

Pedestrian detection is an established instance of computer vision task. Pedestrian detection from the color images has achieved robust performance but in the night time or in bad light conditions it has low detection accuracy. Thermal images are used for detecting people at night time, foggy weather or in bad lighting situations when color images have a lower vision. But in the daytime where the surroundings are warm or warmer than pedestrians then the thermal image has lower accuracy. Hence thermal and color image pair can be a solution but it is expensive to capture color-thermal pair and misaligned imagery can cause low detection accuracy. We proposed a network that achieved better accuracy by extending the prior works which introduced the use of the saliency map in pedestrian detection tasks from the thermal images into instance-level segmentation. We worked on a subdivision of KAIST Multispectral Pedestrian Detection Dataset [8] which has pixel-level annotations. We have trained Mask-RCNN for pedestrian detection task and report the added effect of saliency maps generated using PiCA-Net. We have achieved an accuracy of 88.14% over day and 91.84% over night images. . So, our model has reduced the miss rate by 24.1% and 23% over the existing state-of-the-art method in day and night images.

[...] Read more.
A Comparative Evaluation of Feature Extraction and Similarity Measurement Methods for Content-based Image Retrieval

By S.M. Mohidul Islam Rameswar Debnath

DOI: https://doi.org/10.5815/ijigsp.2020.06.03, Pub. Date: 8 Dec. 2020

Content-based image retrieval is the popular approach for image data searching because in this case, the searching process analyses the actual contents of the image rather than the metadata associated with the image. It is not clear from prior research which feature or which similarity measure performs better among the many available alternatives as well as what are the best combinations of them in content-based image retrieval. We performed a systematic and comprehensive evaluation of several visual feature extraction methods as well as several similarity measurement methods for this case. A feature vector is created after color and/or texture and/or shape features extraction. Then similar images are retrieved using different similarity measures. From experimental results, we found that color moment and wavelet packet entropy features are most effective whereas color autocorrelogram, wavelet moment, and invariant moment features show narrow result. As a similarity measure, cosine and correlation measures are robust in maximum cases; Standardized L2 in few cases and on average, city block measure retrieves more similar images whereas L1 and Mahalanobis measures are less effective in maximum cases. This is the first such system to be informed by a rigorous comparative analysis of the total six features and twelve similarity measures.

[...] Read more.
Variance Analysis Based Mango Recognition Using Correlation Distance

By Farhana Tazmim Pinki S.M. Mohidul Islam

DOI: https://doi.org/10.5815/ijigsp.2020.05.04, Pub. Date: 8 Oct. 2020

Mango plays a major role in the Agro industry and it is a very popular fruit to most of the people due to its flavor and taste. There are many varieties of mangoes that are differentiable based on their various characteristics. Sometimes it is difficult and time consuming for general people or farmers to categorize the mango into different types due to intra-class variation among various types of mangoes. This paper has proposed an automatic system to recognize mangoes thus it becomes convenient to identify various types of mangoes. In this method, mangoes are recognized into different categories based on variance analysis or data dispersion measures. Measures include five number summary, variance, mean deviation, skewness, coefficient of variation which are used as features. From both training and query images, feature vectors are created. Correlation is used to recognize mangoes into various categories. The proposed method shows better result than some existing methods.

[...] Read more.
Edibility Detection of Mushroom Using Ensemble Methods

By Nusrat Jahan Pinky S.M. Mohidul Islam Rafia Sharmin Alice

DOI: https://doi.org/10.5815/ijigsp.2019.04.05, Pub. Date: 8 Apr. 2019

Mushrooms are the most familiar delicious food which is cholesterol free as well as rich in vitamins and minerals. Though nearly 45,000 species of mushrooms have been known throughout the world, most of them are poisonous and few are lethally poisonous. Identifying edible or poisonous mushroom through the naked eye is quite difficult. Even there is no easy rule for edibility identification using machine learning methods that work for all types of data. Our aim is to find a robust method for identifying mushrooms edibility with better performance than existing works. In this paper, three ensemble methods are used to detect the edibility of mushrooms: Bagging, Boosting, and random forest. By using the most significant features, five feature sets are made for making five base models of each ensemble method. The accuracy is measured for ensemble methods using five both fixed feature set-based models and randomly selected feature set based models, for two types of test sets. The result shows that better performance is obtained for methods made of fixed feature sets-based models than randomly selected feature set-based models. The highest accuracy is obtained for the proposed model-based random forest for both test sets.

[...] Read more.
Other Articles