Introduction: Breast cancer is the most common cancer among women. Early and timely detection can increase the effectiveness of more treatment options and lead to more efficient treatment. Mammography is the most standard method for detecting breast cancer, but due to some constraints such as low light intensity, especially in dense breasts, some techniques by means of image processing and artificial intelligence have been developed to detect cancer. These techniques consist of three stages: pre-processing, segmentation and extraction of the tumor areas.
Method: In this paper, using a combination of fuzzy inference and coordinate logic filter, mass candidate regions specified. Then, using thresholding, exact location of masses specified and can be detected in the mammogram.
Results: Because intelligent systems without human errors in diagnosis are common, accuracy and speed of detection proposed method is very high and has a significant coefficient p=0.01 for accurate diagnosis compared with a human detection method.
Conclusion: The experimental results on MIAS database emphasis on the superiority of the proposed method compared with validate existing methods. Diagnosis results achieved high accuracy and precision and margin of less than 8% error for the correct diagnosis.