Abdul Basit

Rice is one of the most important crops in the world because of its vast usage irrespective of age andgender. Machine vision and image processing techniques are widely used to detect diseases in plantleaves. In this paper, a prototype system is developed for rice disease onset detection from images inmonochrome light. Initially, images of leaves of rice plant are acquired in controlled environment. Afterpreprocessing, local binary patterns and local ternary patterns are extracted as features of the image.Support vector machine and k-nearest neighbors are applied as classifier to identify the healthy anddiseased image. Training is completed on 70% of the images while testing is done on the remaining 30%images. Experimental values of the results are 0.94, 0.93, 0.98 and 0.93 for Precision, Sensitivity,Specificity and F1 Score respectively. Overall accuracy of the method is 93.55%. The results showencouraging performance of the proposed method.

Motion Estimation (ME) is a very critical part of any video encoder and different fast ME algorithms are proposed to reduce the overhead of computational cost that was available in encoding system. This paper proposes a Modified Adaptive Predict Hexagon Based Search (MAPHS) algorithm to cope the computational requirements. The proposed algorithm predicts the direction of motion by using adaptive rood shaped predictor and then Hexagonal based Search is applied to refine the search process. The adaptive predictor changes own size predetermined motion vector for each macro block. The ME process is speedup by incorporating thresholding technique, which is primarily beneficial for the videos having low motion activities. Moreover, the proposed algorithm has improved the peak signal to noise ratio (PSNR) by adaptively changing the threshold value according to the motion contents of different videos. The experimental results show that the proposed algorithm performs well on both high definition and common intermediate format videos with respect to ME time and number of search points with acceptable PSNR.

Security is very important and in order to avoid any physical contact, identification of human when they are moving is necessary. Ear biometric is one of the methods by which a person can be identified using surveillance cameras. Various techniques have been proposed to increase the ear based recognition systems. In this work, a feature extraction method for human ear recognition based on wavelet transforms is proposed. The proposed features are approximation coefficients and specific details of level two after applying various types of wavelet transforms. Different wavelet transforms are applied to find the suitable wavelet. Minimum Euclidean distance is used as a matching criterion. Results achieved by the proposed method are promising and can be used in real time ear recognition system.

Automatic detection of breast cancer in mammograms is a challenging task in Computer Aided Diagnosis (CAD) techniques. This paper presents a simple methodology for breast cancer detection in digital mammograms. Proposed methodology consists of three major steps, i.e. segmentation of breast region, removal of pectoral muscle and classification of breast muscle into normal and abnormal tissues. Segmentation of breast muscle was performed by employing Otsus segmentation technique, afterwards removal of pectoral muscle is carried out by canny edge detection and straight line approximation technique. In next step, Gray Level Co-occurrence Matrices (GLCM) was created form which several features were extracted. At the end, SVM classifier was trained to classify breast region into normal and abnormal tissues. Proposed methodology was validated on Mini-MIAS database and results were compared with previously proposed techniques, which shows that proposed technique can be reliably apply for breast cancer detection.

With the development of digital image processing, analysis and modeling techniques, automatic retinal image analysis is emerging as an important screening tool for early detection of ophthalmologic disorders such as diabetic retinopathy and glaucoma. In this paper, a robust method for optic disc detection and extraction of the optic disc boundary is proposed to help in the development of computer-assisted diagnosis and treatment of such ophthalmic disease. The proposed method is based on morphological operations, smoothing filters, and the marker controlled watershed transform. Internal and external markers are used to first modify the gradient magnitude image and then the watershed transformation is applied on this modified gradient magnitude image for boundary extraction. This method has shown significant improvement over existing methods in terms of detection and boundary extraction of the optic disc. The proposed method has optic disc detection success rate of 100%, 100%, 100% and 98.9% for the DRIVE, Shifa, CHASE_DB1, and DIARETDB1 databases, respectively. The optic disc boundary detection achieved an average spatial overlap of 61.88%, 70.96%, 45.61%, and 54.69% for these databases, respectively, which are higher than currents methods.