• Balaji.R
• Giridharan.G
• Narmadha.M

Using small item detection in medical images, medical professionals can concentrate on comprehensive disease evaluation, which is an appealing topic for research. Due to the images' diminished intensity and the distraction of pixel dots, which alter the evaluation of the shape, structure, etc., precise item location and classification must overcome several problems. Real-time circumstances frequently require the ability to identify and categorize minute details in images that have undergone medical intervention. The proposed method was developed using the same criteria for assessing the semantic segmentation of small items in medical images. The system was designed with a focus on adapting the idea to other human organ types, including the lung. Axial CT or PET scans of the lung are considered the major input for the described system. It is discussed how to spot tiny items in CT-PET images, distinguish them from the surrounding tissue, and identify the segmented area as a tumor or nodule. Features are recovered from the preprocessed images following the morphology segmentation, which determines the structural aspects of the tiny object being segmented. Nothing more than the feature points from the segmented pictures produced by morphology and processing make up the feature vectors. These two inputs are loaded into the ensemble convolutional neural network (CNN) to provide the dual classification results.

To Classify the object with sample image in low latency of the image. > To Evaluate the decision making system for nodule or tumor class is determined.

"LUNG NODULE DETECTION USING TWO STAGE CONVOLUTION NEURAL NETWORK WITH KAZE EXTRACTION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 9, page no.f323-f327, September-2023, Available :http://www.jetir.org/papers/JETIR2309539.pdf

"LUNG NODULE DETECTION USING TWO STAGE CONVOLUTION NEURAL NETWORK WITH KAZE EXTRACTION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 9, page no. ppf323-f327, September-2023, Available at : http://www.jetir.org/papers/JETIR2309539.pdf