Identification and diagnosis of periapical lesions, which suggest pathological alterations surrounding the root ends of teeth, depend heavily on periapical radiographs. To help clinicians identify these lesions, precise and effective image segmentation techniques are crucial, mainly when working with big datasets of unlabelled radiography images. This work presents a unique method for separating periapical lesions from unlabelled periapical radiographs employing sophisticated image processing techniques—such as edge identification, contour detection, and enhancement filtering. In order to improve edge details, the suggested methodology starts by applying Gaussian blurring and high-pass filtering on the radiography pictures. Canny edge detection is then used to draw attention to noticeable lesion boundaries after posterization has been used to simplify the image and lower noise. Regions of interest are isolated by extracting contours with an emphasis on locating the lowest points surrounding the root structures. Lastly, these areas are cropped and examined using bounding boxes, making it easier to find lesions. The effectiveness of this segmentation process in separating lesion-like structures is demonstrated by experimental results on unlabelled periapical radiographs, opening the door for additional study and advancement in automated dental diagnostic systems.