Recognizing plants is a vital problem especially for biologists, agricultural researchers, and environmentalists. Plant recognition can be performed by human experts manually but it is a time consuming and low-efficiency process. Automation of plant recognition is an important process for the fields working with plants.
This paper presents an approach for plant recognition using leaf images. In this study, the proponents demonstrated the development of the system that gives users the ability to identify vegetables based on photographs of the leaves taken with a high definition camera.
At the heart of this system is a modernize process of identification, so as to automate the way of identifying the vegetable plants through leaf image and digital image processing. The system used the GaborFilter, Edge Detection, RGB Color and Grayscale Image to acquire the physical parameter of the leaves.
The output parameters are used to compute well documented metrics for the statistical and shape. Base on the study, the following conclusion are drawn: The system can extract the physical parameters from the leaf’s image that will be used in identifying Vegetable`s. From the extracted leaf parameters, the system provides the statistical analysis and general information of the identified leaf. The used algorithm can organize data and information to useful resources to the future researchers.
RELATED LITERATURE AND STUDIES
Interest in digital image processing methods stems from two principal application areas: improvement of pictorial information for human interpretation; and processing of image data for storage, transmission, and representation for autonomous machine perception.
This chapter has several objectives: (1) to define the scope of the field that we call image processing; (2) to give a historical perspective of the origins of this field; (3) to give an idea of the state of the art in image processing by examining some of the principal areas in which it is applied;(4) to discuss briefly the principal approaches used in digital image processing; (5) to give an overview of the components contained in a typical, general-purpose image processing system; and (6) to provide direction to the books and other literature where image processing work normally is reported.
The researchers used Linear Sequential Model as a research design of the study. Each phase has a set of well-defined goals, and the activities within any phase contribute to the satisfaction of that phase’s goals or perhaps a subsequent phase’s goals.
A. System Analysis:
In this phase, the researchers started gathering different vegetable leaves to be used in the study. Also, determining what leaf feature the program should process in leaf recognizing was established.
B. System Design:
The researchers, in this phase, determine how the project will be physically created:
Graphical User Interface (GUI)
Most of the buttons/controls are located in the main form of the system; Open Camera, Capture, Reset, Add New Vegetables and the Leaf Recognition Button. Also, the captured image of the vegetable leaf will show in this form.
Digital Processing techniques help in manipulation of the digital images by using computers. To get over such flaws and to get originality of information, it has to undergo various phases of processing. After capturing the leaf image, the RGB image is firstly resize to 384 x 256 to fasten the process then convert it into a grayscale image.
The code used to convert RGB value of a pixel into its grayscale value is rgb2gray. Then from grayscale, system will use the Gabor filters. It will extract the Gabor features of the image and finally, it will create a column vector, consisting of the image’s Gabor features. Then training and testing of the data will start using (K-NN) k-nearest neighbor classification is often used in classification problems.
D. Testing and Evaluation:
In order to find some bugs that the proposed study might have, researchers administer series of sample leaves if it can acquire the results needed and correctly. The researchers test the accuracy and accuracy of the system by conducting trials of commands. The proponents used 20 trials per leaves in the system has. The accuracy is being tested by how the system correctly responds to the trials conducted by the user. The reliability is tested by how long the system will give correct results as the leaves changes.
This is the stage where the system is first transferred to the users premises and the users get a chance to work with the new system. The system requires Matlab 7.12.0 (R2011a) stable version in order to implement the system with Microsoft Office Access 2003 for database. And also the system needs a high definition camera and a studio-type box with a good lighting. These two will help to give a clear image of the vegetable leaf.
This chapter includes an experiment and data gathering of the program, A Leaf Recognition of Vegetables. With the real-time testing, each of twenty (20) different vegetables leaves undergo in twenty (20) trials. It is to determine the Level of Accuracy and the Reliability of the propose system.
The accuracy of the system tested by the respondent and computed by the proponents is 90.5%. The reliability of the system tested by the respondent and computed by the proponents is 90.75%. And the overall performance of the system, A Leaf Recognition of Vegetable, is 90.625%.
After the processing of data gathered in this study, the researchers have come up with the following conclusion:
- The researchers were able to analyze where the image leaf should undergo.
- The researchers designed the proposed system, A Leaf Recognition of Vegetable which can determine the identification of the leaf tested with the use of Linear Sequential Model as the Research Subject.
- The researchers code or built an image processing system using Matrix Laboratory (Matlab) as the primary programming language.
- The researchers tested the program by experimental evaluation. By having the computed information, the system gives an accuracy of ninety percent (90%). And the researchers also conclude that the propose system can give a reliable result with the percentage of ninety (90%), as long as the vegetable leaf is in the right position, the lighting and the camera device is in good condition, and the shape of the leaf can be distinguish, not deformed nor damaged.
- The researchers can able to implement a useful system for the use of the student and future researchers.
Source: International Journal of Scientific & Technology Research
Authors: Nadine Jaan D. Caldito | Eusebelle B. Dagdagan | Mark G. Estanislao | Kim Leonard B. Jutic | Mary Regina B. Apsay | Marissa G. Chua | Jeffrey F. Calim | Florocito S. Camata