The detection of Lingwu long jujubes in a natural environment is of great significance for robotic picking. Therefore, a lightweight network of target detection based on the SSD (single shot multi-box detector) is presented to meet the requirements of a low computational complexity and enhanced precision. Traditional object detection methods need to load pre-trained weights, cannot change the network structure, and are limited by equipment resource conditions. This study proposes a lightweight SSD object detection method that can achieve a high detection accuracy without loading pre-trained weights and replace the Peleenet network with VGG16 as the trunk, which can acquire additional inputs from all of the previous layers and provide itself characteristic maps to all of the following layers. The coordinate attention module and global attention mechanism are added in the dense block, which boost models to more accurately locate and identify objects of interest. The Inceptionv2 module has been replaced in the first three additional layers of the SSD structure, so the multi-scale structure can enhance the capacity of the model to retrieve the characteristic messages. The output of each additional level is appended to the export of the sub-level through convolution and pooling operations in order to realize the integration of the image feature messages between the various levels. A dataset containing images of the Lingwu long jujubes was generated and augmented using pre-processing techniques such as noise reinforcement, light variation, and image spinning. To compare the performance of the modified SSD model to the original model, a number of experiments were conducted. The results indicate that the mAP (mean average precision) of the modified SSD algorithm for object inspection is 97.32%, the speed of detection is 41.15 fps, and the parameters are compressed to 30.37% of the original networks for the same Lingwu long jujubes datasets without loading pre-trained weights. The improved SSD target detection algorithm realizes a reduction in complexity, which is available for the lightweight adoption to a mobile platform and it provides references for the visual detection of robotic picking.

This paper researches on methods of the color image segmentation method of Lingwu long jujubes based on the maximum entropy to achieve the accuracy of image segmentation and improve accuracy of machine recognition. According to law between the color of Lingwu long jujubes and characteristic of environment, starting from the hue information, this paper is first to explore the difference between the hue of Lingwu long jujubes and the environment which it lives and then use maximum entropy to segment image. It finds optimal threshold by mathematical criterion judging the accuracy of image segmentation. The method of pre-processing of image is mean filter firstly. Then, it extracts hue information of true color image and uses maximum entropy for image segmentation, judging accuracy of image segmentation by segmentation area whether it is in accordance with the 3σ principle. Mathematical morphology is used for smoothing image and eliminating small holes. Finally, segmented image will be obtained through labeling the image by using methods of labeled image and using characteristic parameters for extracting feature. By comparing the segmentation effect with artificial method of the 30 Lingwu long jujubes images, it proves that the color image segmentation method of Lingwu long jujubes based on the maximum entropy has good effect to extract the object region. The accuracy of segmentation rate is up to 89.60%. The time that the algorithm run is 1.3132 s.

No abstract

No abstract