The purpose of this paper is to classify between healthy and sick chicken based on their dropping. Most
chicken farm management system in Malaysia is highly dependent on human surveillance method. This
method, however, does not focus on early disease detection hence, unable to and alert chicken farmers
to take necessary action.. Therefore, the need to improve the biosecurity of chicken poultry production
is essential to prevent infectious disease such as avian influenza. The classification of sick and healthy
chicken based solely on chicken’s excrement using the support vector machine is proposed. First, the
texture is examined using grey-level co-occurrence matrix (GLCM) approach. A GLCM based texture
feature set is derived and used as input for the SVM classifier. Comparison are made using more and
then less extracted features, less extracted features and also applying Gabor filter to these features to
see the effect it has on classification accuracy. Results show that having more features extracted using
GLCM techniques allows for greater classification accuracy.
Nowadays, intelligent vehicles have received a considerable attention among the
researchers to reduce the number of collisions and road accidents. One of the
challenging tasks for these vehicles is road lane detection or road boundaries
detection. In this research, a lane detection algorithm was developed to detect the
right and left lane markers on the road by using two cameras which act as a stereo
vision for the system. It is based on edge detection by using Canny Edge Detection to
reduce unnecessary data on the images and to perform features recognition for the
lane. After the features has been extracted, the algorithm is followed by Hough
Transform method to generate the detected lines on the image obtained from the
stereo vision camera. The algorithm has to work in different environment to be used
in real world applications. The stereo vision algorithm is implemented to generate
disparity map of area. This helps to gain more information on environment, such as the
estimated distance of the lines, the distance of the vehicle to the turns. The experiment
result shows the detection of right and left lane on the road with disparity map to
determine an estimate of the distance of detected lanes from the stereo vision camera.