To address pattern classification problems without involving complex computations, a Hash-Based Convolutional Deep-Thinking Pattern Classifier (HCDTPC) has been developed. This classifier draws inspiration from the human visual system and human thinking logic. The architecture consists of five layers, each serving a distinct purpose. In the initial convolutional layer, informative and distinct features are extracted, mimicking the functionality of the human visual system. The subsequent clustering layer employs a clustering algorithm to group samples within a class into smaller clusters. This arrangement allows the following layer to compare test samples across different clusters, replicating a human-like classification process where the most similar samples should belong to the same group. The hashing layer of the devised system employs a similarity-preserve hashing technique to transform feature maps into hashed data. This transformation reduces the dimensionality of the data while retaining essential similarity information. In the logical layer, Bayesian inference is employed to classify testing samples, mirroring human thought processes. The simulation section assesses the performance of the HCDTPC, demonstrating its effectiveness and efficiency in pattern classification.