In biodiversity research, vast numbers of images produced by scientists provide useful information to many contemporaries.From the (biological) images, elements such as diagnostic hard parts can be used for organism identification at certain level such as genus or species.This results show that the shape characteristic features used are substantial because they can differentiate the specimens based on regions and sexes up to the accuracy of 80% and above.Finally, an application was developed and can be used for the scientific community. Similarly, in biology, images are needed for educational and scientific research purposes.The query image will be compared to the images in the training set and the identification result; normally the system will return the identified species image along with taxon species name but no complete annotations to describe the image.There are a few aspects that are important to consider when developing an automatic identification system, i.e.
This study looks into another invasive process in identification of house shrew () using image analysis and machine learning approaches.
Therefore, image pre-processing is needed to ensure the width, height and pixel size of all images are the same standards. With regards to pattern recognition, features are needed to represent an image, the similarity between two images are then compared using distance function and the similar images to the query image are classified using classifier.
As for query specification, a query image is needed as input whether in query-by-example, query-by-sketch or query-based browsing method.
Both DAISY and SPIDA are generic-based system, which means these systems can be used to recognise many other species.
On the contrary, ABIS and Draw Wing are restricted to insects, which operate by matching specific set of characteristics based on wing venation.
Basically, the identification system is built based on pattern recognition approach.