Monitoring and controlling the weather become essential aspects of mushroom cultivation while considering the effects of temperature, humidity, light intensity, and the amount of carbon dioxide (CO2). The traditional method of mushroom farming is quite challenging due to its lack of control over the weather and cultivation process, resulting in the frequent growth of poisonous mushrooms. This research proposes a framework for a smart testbed for mushroom production with remote-control capabilities and intelligent mechanisms for picking high-quality mushrooms. The proposed method provides real-time insights and automation to optimize mushroom growing techniques by utilizing multimodal sensory observations and the Internet of Things (IoT) technology. We also propose a domain-specific Contrast Limited Adaptive Histogram Equalization (CLAHE) and Laplacian Filter (LF) based DenseNet169 (CLAHE-LF-DenseNet169) model for autonomous mushroom quality determination. Finally, we compare our proposed CLAHE-LF-DenseNet169 model with state-of-the-art techniques, including Resnet50V2 and MobileNet to substantiate the superior performance and efficacy of the model. Experimental outcomes prove that we successfully classify the mushrooms into edible, inedible, and poisonous with an accuracy of 95.21%.