Shitala Prasad

Energy efficient mobile vision system for plant leaf disease identification

Close monitoring, proper control and management of plant diseases are essential in the efficient cultivation of crops. This paper presents a scheme that uses mobile phones for real-time on-field imaging of diseased plants followed by disease diagnosis via analysis of visual phenotypes. A threshold based offloading scheme is employed for judicious sharing of the computational load between the mobile device and a central server at the plant pathology laboratory, thereby offering a trade-off between the power consumption in the mobile device and the transmission cost. The part of the processing carried out in the mobile device includes leaf image segmentation and spotting of disease patch using improved k-means clustering. The algorithm is simple and hence suitable for Android based mobile devices. The segmented image is subsequently communicated to the central server. This ensures reduced transmission cost compared to that in transmitting full leaf image.

An efficient low vision plant leaf shape identification system for smart phones

In computer vision research, the first most important step is to represent the captured object into some mathematical transformed feature vector describing the proper shape, texture and/or color information for the classification. To understand the nature’s biodiversity, together with computer vision (CV), the emerging ubiquitous mobile technologies are now used. Therefore, in this paper, a novel low computational, efficient, and accurate rotation-scale-translation invariant shape profile transform called Angle View Projection (AVP) is proposed. The leaf images captured via mobile devices are transformed to an AVP shape profile curve (a set of four shapelets) and then compacted using Discrete Cosine Transform (DCT) to improve the performance of the system. It also reduces the energy consumption of the device. The algorithm is tested on five different types of leaf datasets: Flavia dataset, 100 plant species leaves dataset, Swedish database, Intelligent Computing Laboratory leaf dataset and Diseased leaf dataset. An ‘Agent’ on mobile device decides whether the module needs to offload to the Server or to compute on the device itself. The experiments carried out clearly indicates that the proposed system outperforms the state-of-the-art with a fast response time even in a low vision environment. AVP also outperforms other methods when tested over incomplete leaves caused due to the physiological or pathological phenomenon. This AVP shape profile based mobile plant biometric system is developed for general applications in our society to better understand the nature and helps in botanical studies and researches.

A wireless dynamic gesture user interface for HCI using hand data glove

In this paper, DG5 hand data glove is used to design an intelligent and efficient human-computer interface to interact with VLC media player. It maps the static keyboard with dynamic human hand gestures with 22 Degree of Freedom (DoF) to interact more natural way with computer. The result is very much appreciated showing the confusion matrix of various gestures used. In this paper, 10 complex gestures are used, that is Play, Pause, Forward, Backward, Next, Previous, Stop, Mute, Full Screen, and Null gestures. To study about the human-hand gestures four different age groups are taken, User A (20-30 years), User B (31-45 years), User C (46-60 years), and User D (61-above years). The decision tree a powerful learning algorithm is used to classify these gestures correctly. This enhances the users interaction level with immersion feeling in augmented reality to 98.88% of accuracy rate.

Mobile augmented reality based interactive teaching & learning system with low computation approach

This paper presents a fast and efficient hand gesture based mobile augmented reality (MAR) system for interactive classroom. It provides a complex visual augmented layer over static slides to understand the concepts more clearly without touching the computer devices or using whiteboard. Simple hand gestures are used to interact with slides while presenting in the classroom or in any conference room with high accuracy and efficiency without any expensive hardware. The gesture path is tracked continuously using a color tracking algorithm proposed. A decision tree is used to make the decisions based on the gestures. The preliminary result indicates that the gesture recognition rate is near about approximately 94% and it is mostly acceptable. This enhances the users interaction level with immersive feeling in immersive environment.

Detection of disease using block-based unsupervised natural plant leaf color image segmentation

A novel unsupervised color image segmentation method is proposed in this paper. First, the image is converted to HSI color model and then it is divided into 5x5 grid matrix resulting into 25 blocks of original leaf image. Each block is then processed separately and passed under an unsupervised segmentation. This segmentation is based on minimizing the energy of each region in the image. This gives the better result in cased of diseased leaf image dataset. This automated system is very much applicable in research work by the botanists specially working with crop diseases and production.

Unsupervised resolution independent based natural plant leaf disease segmentation approach for mobile devices

This paper presents a novel efficient and robust mobile vision system for unsupervised leaf image segmentation in mobile devices which uses a L*a*b* color texture features. In this digital world of ubiquitous computing human machine/mobile interaction (HMI) have packed up its role in human life. A texture based clustering algorithm is developed for plant leaf image segmentation and a pixel wise clustering approach to increase the efficiency. The algorithm is simple and optimized to execute on any Android based mobile devices. The performance of method is evaluated with various types of image resolutions and lighting conditions and results in better accuracy than existing approaches.

Control of computer process using image processing and computer vision for low-processing devices

In this paper a fast and efficient free hand motion detection method is proposed. Presenting an automated intelligent computer vision based HCI system to control and interact without skin-color MAP algorithm to detect motion with more accurate and more natural and efficient way. The experiment involves a very simple mathematics for color tolerance and for motion detection used a trigonometric concept further for action performance based on the gesture definition to compute on low-computing machine such as mobile devices. Here, hand gesture is used to operate presentation by the presenters for easiness while presenting in front of a large gathering and practically sounds good in performance.

Mobile Mixed Reality Based Damage Level Estimation of Diseased Plant Leaf

This paper presents a new dimension for effective cultivation using Mobile Devices (MD) in this ubiquitous digital world. Novel low cost entropy based key frame selection from online mobile-see-through streaming algorithm is proposed. MD is used to monitor the plant leaf disease with much out user interaction and grades them based on the damage level. The mobile mixed reality algorithms are designed to meet mobile limitations of low computation devices such as Smartphones and tables. The system provides interactive powerful mobile interface in farmers pocket to monitor and control the disease attacked on plant leaves. It is easily deployed and used by anyone-anywhere-anytime. The information is augmented on users screen in realistic time. The proposed system is deployed on Android based mobile for the experiments. The performance evaluation of the proposed system is measured in terms of its response time and found to be acceptable.

Plant leaf disease detection using gabor wavelet transform

This paper explores a new dimension of pattern recognition to detect crop diseases based on Gabor Wavelet Transform. The first proposed plant biometric system consist three modules: (1) spot detection using histogram based segmentation, (2) feature extraction using GWT and (3) feature matching with advance machine learning algorithm, SVM. The experimental results on different disease dataset shows that the GWT is effective and robust algorithm for plant disease detection. The accuracy is around 89% in all circumstances. The developed system is very helpful in biology and botanical studies and also used to guide and make aware the Indian farmers about the crop diseases and their natural and chemical controls to improve the production rate.

Vision System for Medicinal Plant Leaf Acquisition and Analysis

Medicinal plant identification is a challenging but very useful task in computer vision (CV). Deep convolutional neural network (CNN) is promisingly used in plant identification as experimentally proved in this paper. It presents a new setup to capture efficiently plant leaves and are used for classification. Secondly, \(l\alpha \beta \) color space is used to improve the performance of CNN in plant species recognition. For this experiment, two different types of datasets are used showing the robustness of our approach.