Breast cancer is a common, potentially fatal disease that not only effects women but can also affect men. Breast cancer is the most common disease affecting women globally, and is the main cause of morbidity and death. Early and accurate detection of this risky disease is very crucial. A timely and precise identification of breast cancer disease can decrease death rate and also can protect people from additional damage. The traditional methods used for identification of breast cancer detection are very expensive in term of time and cost. The goal of this study is to develop a system which can detect the breast cancer accurately and at early stage. The primary objective of this research study is to make use of histopathological images to identify breast cancer correctly and faster. In the proposed research work we have developed a model with name BCDecNet, which comprises twelve learnable layers, i.e., nine convolution layers and three fully connected (FC) layers. The architecture has a total of thirty layers, including one input layer, eight leaky relu (LR) layers, four relu layer, five maximum pooling layers, 6 batch normalization (BN) layers, one cross channel normalization layer and three dropout layers. The proposed work uses image based data taken from Kaggle online repository. The suggested model achieved 97% accuracy, 96% precision, 96% recall and F1 score. Furthermore, the result of proposed model compared with other hybrid approaches used for diagnosis of breast cancer at early stages. Our model achieved satisfactory result then all other approaches use for breast cancer disease detection. Additionally, the proposed BCDecNet model can be generally applied on other medical images datasets for diagnosis of various diseases.